Re: [Samba] Samba performance

[Linda Walsh]

Alan Hodgson wrote:

On Wed, May 25, 2011 at 08:02:56PM -0700, Juan Pablo wrote:

- 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a
switch configured tu use 802.3ad
- 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024
bytes, controller and disk cache enabled, readahead enabled)
- XFS filesystem (created with the following parameters: size=64k -d
su=1024k,sw=7)
- Average file size in the share: 8 MByte
- Gigabit network composed by Cat5E certified cabling and DLink DGS-3427
gigabit switch.


The way Linux does 803.ad is not really how you might expect.


...


It's still not great though. You'd really be better off with a 10Gb/s interface 
out to your switch if you need to guarantee multiple 1Gb/s connections over a 
small number of simultaneous connections.


	Given my experience with bonded ethernet, I'd have to agree.  
	I'm 'just' waiting for the 10Gb prices to come down.  Still a bit 
out of reach for a home network setup.



BTW...

su=1024k?!?   What raid controller are you using?   Usually 64K is usually
recommended for max performance.  But then above you say RAID strip size is
1024bytes?   There is a difference, no?  Which is it?
Either way: a bit off from optimal.


You want to set your log size to 32768b (not 64k; note: 32768b=128k).

For mount options, I have 'swalloc,largeio,logbsize=256k,nobarrier'.

Note, for nobarrier, you *should* have your system on a UPS, and a battery
backup on the RAID controller's cache (LSI controllers have this, others
may as well).

Note, some perf-related options(from my smb.conf) (with host networking 
tuned as well), I have:

aio read size = 65546
aio write size = 65536
max xmit = 66576
min receivefile size = 65536
map acl inherit = Yes
server schannel = no
socket options = TCP_NODELAY IPTOS_LOWDELAY SO_SNDBUF=4194304 SO_RCVBUF=4194304
use sendfile = yes


Note: I'm not sure why my max xmit is > 64k, I probably had a reason
when I set it up -- not even sure if >64k is legal, it might explain why my
read rates are 6MB/s slower than my writes (119MB/s vs. 125MB/s) over Gb lan.

Those are MAX rates to a linear file -- NOT random small reads/writes, BTW
Though I'll regularly see >50MB in random, with >100MB for large files.










--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Linda Walsh]

Juan Pablo wrote:
Thanks a lot for the advice. It will run these tests and try to find meaningfull 
information from them. I will post back results.


Thanks

Juan Pablo


What type of speeds are you expecting?

With a GB network, your limit is 125MB/s.   I get that
with writes, but max out @around 119MB/s on reads due to the 
not being able to have 'overlapping reads'...;-)...


I found to get max performance, I had to adjust the network
params in both linux and windows.

If I'm totally missing some point, I don't get it.

I notice you are trying to use network bonding.  I had problems
getting network bonding to work correctly.

	have you tried sniffing with 'wireshark'?   Maybe look for 
duplicate packets or retries?   To get optimal speeds you need '0 dups'

and '0 retries'...

I've only been able to optimize a single Gb ethernet connection.
A bonded pair -- even direct from server to Win7 of matched Intel dual-port
G-Pro cards gave lower performance than a single wire.

It's odd though, with smbclient -- I'd think that would use
'lo0'  (no?)  I'd think that would get better.   


I noticed in the test below use of 8MB files.  70MB/s would be
a good speed for reading those over the net.  My best raw speeds were using
16-256MB on multi-gig files.   But opening single files ... I'd try
opening them all first, then sending the data, so you are measuring
data perf.

My maximum write perf was done to a file (from windows)
using:
CF="notrunc,nocreat"; OF=direct
"dd if=/dev/zero of='file' bs=16M count=128 oflag=$OF conv=$CF"


Optimizing the network settings on both the linux server and win7 client
gave me another ~20-30%.

I wouldn't trust my testing now, though, as I recently upgraded, and can't
even get nmbd to run...(sigh)...
1 step forward, 3 steps back!






Test typeLocal (dd) Local (smbclient) Window 7
Case1161  101  



   63
Case2122  119  



68

Case1: Read 1000 files 8 MByte each
Case2: 4 processes each reading 1000 files of 8 MByte each

Any idea how can I debug where the bottleneck is or why I get so low numbers 
when reading from Windows?


strace the smbd process with strace -ttT. Network trace.
Look at "netstat -nt" while the test is running. Send/Recv
queues full? Run top, is the CPU fully busy? There's no
silver bullet for performance tuning unfortunately, sorry.

Volker


--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Stan Hoeppner]

On 6/2/2011 2:24 PM, Juan Pablo wrote:
> Hi Stan,
> 
> Thanks for your feedback and suggestions!

You're welcome.  Let's hope they're beneficial.

> The disk subsystem is composed by:
> 
> - 8 WD2002FAEX SATA 2TB hard drives (7200 RPM, 64MB cache, 4.2 ms avg latency)
> - 1 Intel RAID controller RS2BL080 with 512 MB configured with 1 virtual  
> drive 
> 12.7 TB (hardware RAID 5 with 1 MB stripe size, caches enabled,  read-ahead 
> enabled)
> 
> In your experience, should I expect higher performance from this hardware?

That depends on your target workload(s).  You're currently achieving
single stream read performance of 780 MB/s, over 110MB/s per drive.
That's a really good streaming read, close to peak drive read performance.

The problem I see is when you have 4 readers (Win7 clients) reading
4,000 files each.  If these are 16,000 unique files, not each Win7
machine reading the same 4,000 files, i.e. no cache benefit, then I
don't think your disk heads are going be able to seek fast enough to
service all the read requests and hit wire speed SMB.  If your
production load will be significantly less than this artificial test
load, you may be fine.

> Will try the ramdisk test you are suggesting and post back the results. 
> Thanks 
> for the suggestion!

The results should be informative, one way or the other.

> I have jumbo frames enabled in the switches but windows drivers for the Intel 
> network cards don't have the option to enable jumbo frames. I also tried 
> raising 
> the MTU in the linux server but performance was even worse (I thought this 
> was 
> related to the windows NIC driver not supporting MTUs larger than 1500).

Lack of jumbo frames is probably hurting your wire performance due to
increased interrupt processing and other factors.  I'm surprised some
Intel NICs don't support jumbo frames.  Must be desktop adapters.  Can
you post the model# of the NICs in the Win7 PCs and those in the server
so I can do some research?

> I also modified windows registry to manually enable smb2 protocol because it 
> was 
> not negotiating smb2. Do you think of any other optimization that can be done 
> on 
> the windows terminals?

I have no experience yet with SMB2 or Win7 so I can't really say.  You
should be able to tune that server and the clients to hit near wire
speed with regular SMB.  I suggest solving that problem first, then
worry about SMB2.

-- 
Stan
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Hi Stan,

Thanks for your feedback and suggestions!


The disk subsystem is composed by:

- 8 WD2002FAEX SATA 2TB hard drives (7200 RPM, 64MB cache, 4.2 ms avg latency)
- 1 Intel RAID controller RS2BL080 with 512 MB configured with 1 virtual  drive 
12.7 TB (hardware RAID 5 with 1 MB stripe size, caches enabled,  read-ahead 
enabled)

In your experience, should I expect higher performance from this hardware?

Will try the ramdisk test you are suggesting and post back the results. Thanks 
for the suggestion!

I have jumbo frames enabled in the switches but windows drivers for the Intel 
network cards don't have the option to enable jumbo frames. I also tried 
raising 
the MTU in the linux server but performance was even worse (I thought this was 
related to the windows NIC driver not supporting MTUs larger than 1500).

I also modified windows registry to manually enable smb2 protocol because it 
was 
not negotiating smb2. Do you think of any other optimization that can be done 
on 
the windows terminals?

Thanks

Juan Pablo




From: Stan Hoeppner 
To: Juan Pablo 
Cc: samba@lists.samba.org
Sent: Thu, June 2, 2011 8:50:21 AM
Subject: Re: [Samba] Samba performance

On 5/25/2011 10:02 PM, Juan Pablo wrote:

> OS access: 
> Simultaneous read (4 processes): 118 MByte/s average

> Samba local access:
> Simultaneous read (4 processes): 102 MByte/s average

> Samba server from Windows 7:
> Simultaneous read (4 terminals):  70 MByte/s average

The first two results above demonstrate a slow disk subsystem not
suitable for streaming multiple files to multiple concurrent clients at
high data rates.  Your spindles are too slow and/or you don't have
enough to satisfy your test methodology.  Four concurrent dd copies
yields 118 MB/s per process, only ~15% disk headroom above wire speed
GbE.  Your smbd+smbclient local process disk bandwidth overhead appears
to be roughly 13 percent.  I don't know what the optimal percent here
should be but 13% above a dd copy process seems reasonable given the
additional data movement through smbd and smbclient buffers.

It is clear that you don't have enough head seek performance for 4 or
more client streams of 1000 x 8MB files.  This doesn't necessarily
address the 30% drop in over the wire to Win7 client performance, but
we'll get to that later.  To confirm the disk deficiency issue, I
recommend the following test:

Make a 2GB tmpfs ramdisk on the server and run your tests against it,
albeit with 200 instead of 1000 8MB files.  Instructions:
http://prefetch.net/blog/index.php/2006/11/30/creating-a-ramdisk-with-linux/

This will tell you if your server block storage subsystem is part of the
problem, and will give you a maximum throughput per Samba process
baseline.  You should get something like 5GB/s+ local smbclient
throughput from a tmpfs ramdisk on that Xeon platform with its raw
25GB/s memory bandwidth.

Run a single Win7 workstation SMB test copy to a freshly booted machine
so most of the memory is free for buffering the inbound files.  This
will mostly eliminate the slow local disk as a bottleneck.

Now run your 4 concurrent Win7 client test and compare to the single
client test results.  This should tell you if you have a bonding problem
or not, either in the server NICs or the switch.

You didn't mention jumbo frames.  Enable jumbo if not already.  It may help.

Something else to consider is that the kernel shipped with CentOS 5.6,
2.6.18, the "Pirate" kernel, is now 4.5 years old, released in Sept of
2006 (http://kerneltrap.org/node/7144).  There have been just a few
performance enhancements between 2.6.18 and 3.0, specifically to the
network stack. ;)  The CentOS packages are older than dirt as well.  If
you're not wed to CentOS you should look at more recent distros.

-- 
Stan
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Thanks a lot for the advice. It will run these tests and try to find 
meaningfull 
information from them. I will post back results.

Thanks

Juan Pablo





From: Volker Lendecke 
To: Juan Pablo 
Cc: Jeremy Allison ; samba@lists.samba.org
Sent: Thu, June 2, 2011 3:49:17 AM
Subject: Re: [Samba] Samba performance

On Wed, Jun 01, 2011 at 06:46:51PM -0700, Juan Pablo wrote:
> Hi Volker,
> 
> I've removed the SO_RCVBUF=65536 SO_SNDBUF=65536 and the 3 other setting, 
> reloaded samba and repeated the tests but still getting the same results for 
>the 
>
> local tests and also from Windows.
> 
> I am getting the following results in MBytes/s:
> 
> Test typeLocal (dd) Local (smbclient) Window 7
> Case1161  101 
>  

>63
> Case2122  119 
>  

> 68
> 
> Case1: Read 1000 files 8 MByte each
> Case2: 4 processes each reading 1000 files of 8 MByte each
> 
> Any idea how can I debug where the bottleneck is or why I get so low numbers 
> when reading from Windows?

strace the smbd process with strace -ttT. Network trace.
Look at "netstat -nt" while the test is running. Send/Recv
queues full? Run top, is the CPU fully busy? There's no
silver bullet for performance tuning unfortunately, sorry.

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Stan Hoeppner]

On 5/25/2011 10:02 PM, Juan Pablo wrote:

> OS access: 
> Simultaneous read (4 processes): 118 MByte/s average

> Samba local access:
> Simultaneous read (4 processes): 102 MByte/s average

> Samba server from Windows 7:
> Simultaneous read (4 terminals):  70 MByte/s average

The first two results above demonstrate a slow disk subsystem not
suitable for streaming multiple files to multiple concurrent clients at
high data rates.  Your spindles are too slow and/or you don't have
enough to satisfy your test methodology.  Four concurrent dd copies
yields 118 MB/s per process, only ~15% disk headroom above wire speed
GbE.  Your smbd+smbclient local process disk bandwidth overhead appears
to be roughly 13 percent.  I don't know what the optimal percent here
should be but 13% above a dd copy process seems reasonable given the
additional data movement through smbd and smbclient buffers.

It is clear that you don't have enough head seek performance for 4 or
more client streams of 1000 x 8MB files.  This doesn't necessarily
address the 30% drop in over the wire to Win7 client performance, but
we'll get to that later.  To confirm the disk deficiency issue, I
recommend the following test:

Make a 2GB tmpfs ramdisk on the server and run your tests against it,
albeit with 200 instead of 1000 8MB files.  Instructions:
http://prefetch.net/blog/index.php/2006/11/30/creating-a-ramdisk-with-linux/

This will tell you if your server block storage subsystem is part of the
problem, and will give you a maximum throughput per Samba process
baseline.  You should get something like 5GB/s+ local smbclient
throughput from a tmpfs ramdisk on that Xeon platform with its raw
25GB/s memory bandwidth.

Run a single Win7 workstation SMB test copy to a freshly booted machine
so most of the memory is free for buffering the inbound files.  This
will mostly eliminate the slow local disk as a bottleneck.

Now run your 4 concurrent Win7 client test and compare to the single
client test results.  This should tell you if you have a bonding problem
or not, either in the server NICs or the switch.

You didn't mention jumbo frames.  Enable jumbo if not already.  It may help.

Something else to consider is that the kernel shipped with CentOS 5.6,
2.6.18, the "Pirate" kernel, is now 4.5 years old, released in Sept of
2006 (http://kerneltrap.org/node/7144).  There have been just a few
performance enhancements between 2.6.18 and 3.0, specifically to the
network stack. ;)  The CentOS packages are older than dirt as well.  If
you're not wed to CentOS you should look at more recent distros.

-- 
Stan


-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Volker Lendecke]

On Wed, Jun 01, 2011 at 06:46:51PM -0700, Juan Pablo wrote:
> Hi Volker,
> 
> I've removed the SO_RCVBUF=65536 SO_SNDBUF=65536 and the 3 other setting, 
> reloaded samba and repeated the tests but still getting the same results for 
> the 
> local tests and also from Windows.
> 
> I am getting the following results in MBytes/s:
> 
> Test typeLocal (dd) Local (smbclient) Window 7
> Case1161  101 
>  
>63
> Case2122  119 
>   
> 68
> 
> Case1: Read 1000 files 8 MByte each
> Case2: 4 processes each reading 1000 files of 8 MByte each
> 
> Any idea how can I debug where the bottleneck is or why I get so low numbers 
> when reading from Windows?

strace the smbd process with strace -ttT. Network trace.
Look at "netstat -nt" while the test is running. Send/Recv
queues full? Run top, is the CPU fully busy? There's no
silver bullet for performance tuning unfortunately, sorry.

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Hi Volker,

I've removed the SO_RCVBUF=65536 SO_SNDBUF=65536 and the 3 other setting, 
reloaded samba and repeated the tests but still getting the same results for 
the 
local tests and also from Windows.

I am getting the following results in MBytes/s:

Test typeLocal (dd) Local (smbclient) Window 7
Case1161  101  
   63
Case2122  119   
68

Case1: Read 1000 files 8 MByte each
Case2: 4 processes each reading 1000 files of 8 MByte each

Any idea how can I debug where the bottleneck is or why I get so low numbers 
when reading from Windows?

Thanks


Juan Pablo





From: Volker Lendecke 
To: Juan Pablo 
Cc: Jeremy Allison ;  samba@lists.samba.org
Sent: Fri, May 27, 2011 11:25:31 AM
Subject: Re: [Samba] Samba performance

On Fri, May 27, 2011 at 06:34:50AM -0700, Juan Pablo wrote:
> Hi Volker,
> 
> I am using the following socket options:
> 
> socket options = TCP_NODELAY IPTOS_LOWDELAY SO_RCVBUF=65536 SO_SNDBUF=65536

Just remove the SO_RCVBUF=65536 SO_SNDBUF=65536 settings.
Unless you're on a very old Linux or other Unix the kernel
is far better off figuring out that itself.

> read raw = yes
> write raw = yes
> max xmit = 65535

Just remove these 3 settings. If it's still slow after that,
we need to do more analysis.

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Volker Lendecke]

On Fri, May 27, 2011 at 06:34:50AM -0700, Juan Pablo wrote:
> Hi Volker,
> 
> I am using the following socket options:
> 
> socket options = TCP_NODELAY IPTOS_LOWDELAY SO_RCVBUF=65536 SO_SNDBUF=65536

Just remove the SO_RCVBUF=65536 SO_SNDBUF=65536 settings.
Unless you're on a very old Linux or other Unix the kernel
is far better off figuring out that itself.

> read raw = yes
> write raw = yes
> max xmit = 65535

Just remove these 3 settings. If it's still slow after that,
we need to do more analysis.

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Hi Alan,

My setup is similar to what you are suggesting: options bond0 miimon=100 mode=4 
lacp_rate=1

I am still not happy with how bonding is working. It seems that the first 3 
connections are getting assigned to different links but the 4th is using one of 
the previously used ones. I will add xmit_hash_policy=layer2+3 to see if I get 
any improvement.

I thought the algorithm was smart enough to get the first n connections (being 
n 
the number of available links) into different links. 


Thanks for your advice! I will try your suggestion.

Regards

Juan Pablo


From: Alan Hodgson 
To: samba@lists.samba.org
Sent: Thu, May 26, 2011 2:45:12 PM
Subject: Re: [Samba] Samba performance

On Wed, May 25, 2011 at 08:02:56PM -0700, Juan Pablo wrote:
> > - 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a
> > switch configured tu use 802.3ad
> > - 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024
> > bytes, controller and disk cache enabled, readahead enabled)
> > - XFS filesystem (created with the following parameters: size=64k -d
> > su=1024k,sw=7)
> > - Average file size in the share: 8 MByte
> > - Gigabit network composed by Cat5E certified cabling and DLink DGS-3427
> > gigabit switch.

The way Linux does 803.ad is not really how you might expect.

Basically, it isn't really smart about how it assigns individual network 
streams to ports. It hashes destination MAC addresses and then randomly 
assigns them to ports. It doesn't pay attention to current traffic levels. It 
is 

quite likely that all 4 connections ended up using the same port - which you 
could see by monitoring the total packets transmitted on each interface during 
your test.

You can make it somewhat more likely to use more ports by using a different 
hashing algorithm. I've had the best results with something like this in 
modprobe.conf:

"options bonding mode=802.3ad miimon=100 lacp_rate=1 
xmit_hash_policy=layer2+3"

It's still not great though. You'd really be better off with a 10Gb/s interface 
out to your switch if you need to guarantee multiple 1Gb/s connections over a 
small number of simultaneous connections.

-- 
The whole universe is change and life itself is but what you deem it.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Hi Daniel,

Thanks for your advice.

I thought that XFS was a good alternative as I have plenty of unused CPU. I've 
been googling and found that XFS had good performance but used more CPU than 
other alternatives.

I will like to get as much as I can from this fileserver. I will try EXT4 to 
see 
if I can make any improvement with it but I am currently seeing a very big 
difference on the average speed I get from reading directly (using the dd 
command) to what I get with smbclient without passing through the switch.

Thanks again! 

Juan Pablo





From: Daniel Deptuła 
To: samba@lists.samba.org
Cc: jhur...@yahoo.com
Sent: Thu, May 26, 2011 1:19:03 PM
Subject: Re: [Samba] Samba performance

W dniu 2011-05-26 05:02, Juan Pablo pisze:
> Hi everyone,
>
> I'm trying to use samba in a small video post production house but we are not
> getting the performance we expected.
>
> Our setup:
>
> - CenOS 5.6 x86-64
> - samba.x86_64 (3.0.33-3.29.el5_6.2 and 3.6.0rc1)
> - Intel based server (One 4 core Xeon E5620  @ 2.40GHz, 8 GB RAM)
> - 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a 
switch
> configured tu use 802.3ad
> - 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024 bytes,
> controller and disk cache enabled, readahead enabled)
> - XFS filesystem (created with the following parameters: size=64k -d
> su=1024k,sw=7)
> - Average file size in the share: 8 MByte
> - Gigabit network composed by Cat5E certified cabling and DLink DGS-3427 
>gigabit
> switch.
> - Intel I7 based terminals with Intel gigabit NIC, running Windows 7
>
>
> Test results:
>
> OS access:
>
> Sequential write (1 x 31 GByte file): 500 MByte/s
> Sequential read (1 x 31 GByte file): 780 MByte/s
> Write (1000 files 8 MByte each): 249 MByte/s average
> Read (1000 files 8 MByte each): 158 MByte/s average
> Simultaneous write (4 processes each writing 1000 files of 8 MByte each ): 188
> MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 118
> MByte/s average
>
> Samba local access (stock CentOS samba 3.0.33 connecting from the same server
> with smbclient):
>
> Sequential read (1 x 31 GByte file):  267 MByte/s
> Read (1000 files 8 MByte each): 71 MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 102
> MByte/s average
>
> Samba local access (Samba 3.6.0rc1 compiled from GIT repo. Connecting from the
> same server with smbclient):
>
> Read (1000 files 8 MByte each): 95 MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 103
> MByte/s average
>
> Samba server accessed from Windows 7 terminals (samba 3.6.0rc1):
>
> Read (1 terminal copying from samba fileserver to local disk 1000 files 8 
MByte
> each): 60 MByte/s average
> Simultaneous read (4 terminals each copying from samba fileserver to local 
disk
> 1000 files of 8 MByte each): 70 MByte/s average
>
> Note: Simultaneos read speed is measured adding the size of all transfered 
>files
> and dividing it by the time taken to transfer these files.
>
> I will appreciate any feedback about the results we are getting and advice on
> how to improve this.
>
> Thanks in advance
>
> Juan Pablo
Maybe try the ext4 filesystem? With a new kernel - with stable support 
for it. Many tests have shown that ext4 is faster than XFS, but also 
remember to tune the parameters when creating the filesystem. You can 
try several different configurations and compare their performance 
(performance for the same parameters can be different on different 
hardware and RAID configurations, so options recommended by other people 
are not always the best for you). Filesystem mount options are also 
important!

The second thing is network - some switches do not do port trunking well 
- for example they use always use one wire even if there are 2 or more 
connected in a trunk - so it does not improve performance - only the 
reliability. Usually also one data stream does not go through more than 
one wire, so the only possibility to get 4 Gbit speed from your server 
is to connect 4 simultaneously downloading stations to the switch. You 
can check the bandwidth usage on each interface of the server with the 
iftop command. For measuring the network performance I recommend also 
the iperf tool.

Also google about network and tcp tuning in linux (parameters like 
txqueuelen, buffer sizes etc).

About tuning samba performance you can read for example here:
http://www.samba.org/samba/docs/man/Samba-HOWTO-Collection/speed.html
But also in many other places on the Internet.

Best regards,
Daniel
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Hi Volker,

I am using the following socket options:

socket options = TCP_NODELAY IPTOS_LOWDELAY SO_RCVBUF=65536 SO_SNDBUF=65536

I've been testing with lower rcvbuf and sndbuf values but this seems to be the 
best. Any advice on how to calculate this value besides trial and error?

Following is my smb.conf:

[global]
workgroup = Workgrouplasp
encrypt passwords = yes
security = user
smb passwd file = /etc/samba/smbpasswd
netbios name = LASPFS
server string = LASP Linux File Server
dns proxy = no
log level = 1
oplocks = yes
level2 oplocks = yes
socket options = TCP_NODELAY IPTOS_LOWDELAY SO_RCVBUF=65536 
SO_SNDBUF=65536
guest account = samba
map to guest = Bad User
map to guest = Bad Password
read raw = yes
write raw = yes
max xmit = 65535
dead time = 15
getwd cache = yes
interfaces = 192.168.1.5/255.255.255.0
bind interfaces only = yes
socket address = 192.168.1.5
max protocol = SMB2

[laspfiles] 
comment = Public File Storage
browseable = yes
force user = samba
force group = samba
path = /srv/samba/filestore/laspfiles
read only = no
guest ok = yes

My focus was to tune the fileserver for several 8MB file transfer. Any idea on 
how to improve this?

Thanks for your help!

Juan Pablo





From: Volker Lendecke 
To: Juan Pablo 
Cc: Jeremy Allison ; samba@lists.samba.org
Sent: Thu, May 26, 2011 2:27:20 PM
Subject: Re: [Samba] Samba performance

On Thu, May 26, 2011 at 10:14:31AM -0700, Juan Pablo wrote:
> Hi Jeremy,
> 
> Thanks for your reply!
> 
> The tests we did with the Windows 7 terminals was using smb2. 
> 
> When we enabled smb2 in samba we saw in samba logs that it
> was not being  used.  We modified Windows 7 registry as
> described in
>http://www.techemperor.com/2009/09/21/manual-patch-for-windows-vistaserver-2008-smb2-flaw/
>/
> to enable smb2 to start using smb2. Once this was done
> smb2 was  negotiated but there was no speed difference.
> 
> The OS read test is done iterating from 0 to 999 a dd
> if=testFile-xxx  of=/dev/null bs=1k. The samba local
> access test is done with smbclient  from the same machine
> sending the output to /dev/null. Is the speed decrease
> (from 158 MB/s to 71 MB/s) from what I get when I test
> from  the OS to what I get with samba normal? 

With smbclient you should get near wire speed. Do you have
"socket options" set in your smb.conf?

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Alan Hodgson]

On Wed, May 25, 2011 at 08:02:56PM -0700, Juan Pablo wrote:
> > - 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a
> > switch configured tu use 802.3ad
> > - 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024
> > bytes, controller and disk cache enabled, readahead enabled)
> > - XFS filesystem (created with the following parameters: size=64k -d
> > su=1024k,sw=7)
> > - Average file size in the share: 8 MByte
> > - Gigabit network composed by Cat5E certified cabling and DLink DGS-3427
> > gigabit switch.

The way Linux does 803.ad is not really how you might expect.

Basically, it isn't really smart about how it assigns individual network 
streams to ports. It hashes destination MAC addresses and then randomly 
assigns them to ports. It doesn't pay attention to current traffic levels. It 
is 
quite likely that all 4 connections ended up using the same port - which you 
could see by monitoring the total packets transmitted on each interface during 
your test.

You can make it somewhat more likely to use more ports by using a different 
hashing algorithm. I've had the best results with something like this in 
modprobe.conf:

"options bonding mode=802.3ad miimon=100 lacp_rate=1 
xmit_hash_policy=layer2+3"

It's still not great though. You'd really be better off with a 10Gb/s interface 
out to your switch if you need to guarantee multiple 1Gb/s connections over a 
small number of simultaneous connections.

-- 
The whole universe is change and life itself is but what you deem it.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Volker Lendecke]

On Thu, May 26, 2011 at 10:14:31AM -0700, Juan Pablo wrote:
> Hi Jeremy,
> 
> Thanks for your reply!
> 
> The tests we did with the Windows 7 terminals was using smb2. 
> 
> When we enabled smb2 in samba we saw in samba logs that it
> was not being  used.  We modified Windows 7 registry as
> described in
> http://www.techemperor.com/2009/09/21/manual-patch-for-windows-vistaserver-2008-smb2-flaw/
> to enable smb2 to start using smb2. Once this was done
> smb2 was  negotiated but there was no speed difference.
> 
> The OS read test is done iterating from 0 to 999 a dd
> if=testFile-xxx  of=/dev/null bs=1k. The samba local
> access test is done with smbclient  from the same machine
> sending the output to /dev/null. Is the speed decrease
> (from 158 MB/s to 71 MB/s) from what I get when I test
> from  the OS to what I get with samba normal? 

With smbclient you should get near wire speed. Do you have
"socket options" set in your smb.conf?

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Juan Pablo]

Hi Jeremy,

Thanks for your reply!

The tests we did with the Windows 7 terminals was using smb2. 

When we enabled smb2 in samba we saw in samba logs that it was not being  used. 
We modified Windows 7 registry as described in 
http://www.techemperor.com/2009/09/21/manual-patch-for-windows-vistaserver-2008-smb2-flaw/
 to enable smb2 to start using smb2. Once this was done smb2 was  negotiated 
but 
there was no speed difference.

The OS read test is done iterating from 0 to 999 a dd if=testFile-xxx  
of=/dev/null bs=1k. The samba local access test is done with smbclient  from 
the 
same machine sending the output to /dev/null. Is the speed  decrease  (from 158 
MB/s to 71 MB/s) from what I get when I test from  the OS to what I get with 
samba normal? 


Juan Pablo




From: Jeremy Allison 
To: Juan Pablo 
Cc: samba@lists.samba.org
Sent: Thu, May 26, 2011 1:16:02 PM
Subject: Re: [Samba] Samba performance

On Wed, May 25, 2011 at 08:02:56PM -0700, Juan Pablo wrote:
> Hi everyone,
> 
> I'm trying to use samba in a small video post production house but we are not 
> getting the performance we expected.
> 
> Our setup:
> 
> - CenOS 5.6 x86-64
> - samba.x86_64 (3.0.33-3.29.el5_6.2 and 3.6.0rc1)
> - Intel based server (One 4 core Xeon E5620  @ 2.40GHz, 8 GB RAM)
> - 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a 
> switch 
>
> configured tu use 802.3ad
> - 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024 bytes, 
> controller and disk cache enabled, readahead enabled)
> - XFS filesystem (created with the following parameters: size=64k -d 
> su=1024k,sw=7)
> - Average file size in the share: 8 MByte
> - Gigabit network composed by Cat5E certified cabling and DLink DGS-3427 
>gigabit 
>
> switch.
> - Intel I7 based terminals with Intel gigabit NIC, running Windows 7
> 
> 
> Test results:
> 
> OS access: 
> 
> Sequential write (1 x 31 GByte file): 500 MByte/s
> Sequential read (1 x 31 GByte file): 780 MByte/s
> Write (1000 files 8 MByte each): 249 MByte/s average
> Read (1000 files 8 MByte each): 158 MByte/s average
> Simultaneous write (4 processes each writing 1000 files of 8 MByte each ): 
> 188 

> MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 118 
> MByte/s average
> 
> Samba local access (stock CentOS samba 3.0.33 connecting from the same server 
> with smbclient):
> 
> Sequential read (1 x 31 GByte file):  267 MByte/s
> Read (1000 files 8 MByte each): 71 MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 102 
> MByte/s average
> 
> Samba local access (Samba 3.6.0rc1 compiled from GIT repo. Connecting from 
> the 

> same server with smbclient):
> 
> Read (1000 files 8 MByte each): 95 MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 103 
> MByte/s average
> 
> Samba server accessed from Windows 7 terminals (samba 3.6.0rc1):
> 
> Read (1 terminal copying from samba fileserver to local disk 1000 files 8 
> MByte 
>
> each): 60 MByte/s average
> Simultaneous read (4 terminals each copying from samba fileserver to local 
> disk 
>
> 1000 files of 8 MByte each): 70 MByte/s average
> 
> Note: Simultaneos read speed is measured adding the size of all transfered 
>files 
>
> and dividing it by the time taken to transfer these files.
> 
> I will appreciate any feedback about the results we are getting and advice on 
> how to improve this.

If you're using 3.6.0 and Windows 7 clients try turning on SMB2 support
by setting "max protocol = smb2" in the [global] section of your smb.conf.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Volker Lendecke]

On Thu, May 26, 2011 at 09:16:02AM -0700, Jeremy Allison wrote:
> If you're using 3.6.0 and Windows 7 clients try turning on SMB2 support
> by setting "max protocol = smb2" in the [global] section of your smb.conf.

Well, using smbclient should definitely get better
performance. Something is wrong here...

Volker

-- 
SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen
phone: +49-551-37-0, fax: +49-551-37-9
AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Daniel Deptuła]

W dniu 2011-05-26 05:02, Juan Pablo pisze:

Hi everyone,

I'm trying to use samba in a small video post production house but we are not
getting the performance we expected.

Our setup:

- CenOS 5.6 x86-64
- samba.x86_64 (3.0.33-3.29.el5_6.2 and 3.6.0rc1)
- Intel based server (One 4 core Xeon E5620  @ 2.40GHz, 8 GB RAM)
- 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a switch
configured tu use 802.3ad
- 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024 bytes,
controller and disk cache enabled, readahead enabled)
- XFS filesystem (created with the following parameters: size=64k -d
su=1024k,sw=7)
- Average file size in the share: 8 MByte
- Gigabit network composed by Cat5E certified cabling and DLink DGS-3427 gigabit
switch.
- Intel I7 based terminals with Intel gigabit NIC, running Windows 7


Test results:

OS access:

Sequential write (1 x 31 GByte file): 500 MByte/s
Sequential read (1 x 31 GByte file): 780 MByte/s
Write (1000 files 8 MByte each): 249 MByte/s average
Read (1000 files 8 MByte each): 158 MByte/s average
Simultaneous write (4 processes each writing 1000 files of 8 MByte each ): 188
MByte/s average
Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 118
MByte/s average

Samba local access (stock CentOS samba 3.0.33 connecting from the same server
with smbclient):

Sequential read (1 x 31 GByte file):  267 MByte/s
Read (1000 files 8 MByte each): 71 MByte/s average
Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 102
MByte/s average

Samba local access (Samba 3.6.0rc1 compiled from GIT repo. Connecting from the
same server with smbclient):

Read (1000 files 8 MByte each): 95 MByte/s average
Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 103
MByte/s average

Samba server accessed from Windows 7 terminals (samba 3.6.0rc1):

Read (1 terminal copying from samba fileserver to local disk 1000 files 8 MByte
each): 60 MByte/s average
Simultaneous read (4 terminals each copying from samba fileserver to local disk
1000 files of 8 MByte each): 70 MByte/s average

Note: Simultaneos read speed is measured adding the size of all transfered files
and dividing it by the time taken to transfer these files.

I will appreciate any feedback about the results we are getting and advice on
how to improve this.

Thanks in advance

Juan Pablo
Maybe try the ext4 filesystem? With a new kernel - with stable support 
for it. Many tests have shown that ext4 is faster than XFS, but also 
remember to tune the parameters when creating the filesystem. You can 
try several different configurations and compare their performance 
(performance for the same parameters can be different on different 
hardware and RAID configurations, so options recommended by other people 
are not always the best for you). Filesystem mount options are also 
important!


The second thing is network - some switches do not do port trunking well 
- for example they use always use one wire even if there are 2 or more 
connected in a trunk - so it does not improve performance - only the 
reliability. Usually also one data stream does not go through more than 
one wire, so the only possibility to get 4 Gbit speed from your server 
is to connect 4 simultaneously downloading stations to the switch. You 
can check the bandwidth usage on each interface of the server with the 
iftop command. For measuring the network performance I recommend also 
the iperf tool.


Also google about network and tcp tuning in linux (parameters like 
txqueuelen, buffer sizes etc).


About tuning samba performance you can read for example here:
http://www.samba.org/samba/docs/man/Samba-HOWTO-Collection/speed.html
But also in many other places on the Internet.

Best regards,
Daniel
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance

[Jeremy Allison]

On Wed, May 25, 2011 at 08:02:56PM -0700, Juan Pablo wrote:
> Hi everyone,
> 
> I'm trying to use samba in a small video post production house but we are not 
> getting the performance we expected.
> 
> Our setup:
> 
> - CenOS 5.6 x86-64
> - samba.x86_64 (3.0.33-3.29.el5_6.2 and 3.6.0rc1)
> - Intel based server (One 4 core Xeon E5620  @ 2.40GHz, 8 GB RAM)
> - 4 Intel Gigagit ethernet NIC ports with 802.3ad bonding connected to a 
> switch 
> configured tu use 802.3ad
> - 8 2TB 7.2 krpm SATA disks with hardware RAID5 (RAID stripe size 1024 bytes, 
> controller and disk cache enabled, readahead enabled)
> - XFS filesystem (created with the following parameters: size=64k -d 
> su=1024k,sw=7)
> - Average file size in the share: 8 MByte
> - Gigabit network composed by Cat5E certified cabling and DLink DGS-3427 
> gigabit 
> switch.
> - Intel I7 based terminals with Intel gigabit NIC, running Windows 7
> 
> 
> Test results:
> 
> OS access: 
> 
> Sequential write (1 x 31 GByte file): 500 MByte/s
> Sequential read (1 x 31 GByte file): 780 MByte/s
> Write (1000 files 8 MByte each): 249 MByte/s average
> Read (1000 files 8 MByte each): 158 MByte/s average
> Simultaneous write (4 processes each writing 1000 files of 8 MByte each ): 
> 188 
> MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 118 
> MByte/s average
> 
> Samba local access (stock CentOS samba 3.0.33 connecting from the same server 
> with smbclient):
> 
> Sequential read (1 x 31 GByte file):  267 MByte/s
> Read (1000 files 8 MByte each): 71 MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 102 
> MByte/s average
> 
> Samba local access (Samba 3.6.0rc1 compiled from GIT repo. Connecting from 
> the 
> same server with smbclient):
> 
> Read (1000 files 8 MByte each): 95 MByte/s average
> Simultaneous read (4 processes each reading 1000 files of 8 MByte each): 103 
> MByte/s average
> 
> Samba server accessed from Windows 7 terminals (samba 3.6.0rc1):
> 
> Read (1 terminal copying from samba fileserver to local disk 1000 files 8 
> MByte 
> each): 60 MByte/s average
> Simultaneous read (4 terminals each copying from samba fileserver to local 
> disk 
> 1000 files of 8 MByte each): 70 MByte/s average
> 
> Note: Simultaneos read speed is measured adding the size of all transfered 
> files 
> and dividing it by the time taken to transfer these files.
> 
> I will appreciate any feedback about the results we are getting and advice on 
> how to improve this.

If you're using 3.6.0 and Windows 7 clients try turning on SMB2 support
by setting "max protocol = smb2" in the [global] section of your smb.conf.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] samba performance multi-thread and multi core

[John Drescher]

> I have question about samba performance with  multi-thread and multi core
> cpu.
>
> What can we do for samba performance with  multi-thread and multi core ?
>
Each connected user gets their own process and thus threads. The
system will balance the threads over the cpus.

John
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/options/samba

Re: [Samba] Samba performance issue

[ales-76]

Hi Fabien,

I haven't tried FUSE, and to be honest I will not. I doubt that the performace 
will even match cifs or smbfs implementation. I would say that file systems in 
user space are good for development or special applications, not for general 
purpose. Anyway I you are to try it out please let me know what you find.

Thank you

Ales



 Původní zpráva 
Od: Fabien 
Předmět: Re: [Samba] Samba performance issue
Datum: 11.1.2009 00:29:18

Hello,

as you say, I also think it would be nice to mention the issue in the
documentation to avoid people turning crazy !

If you need the filesystem to be mounted, did you try the fuse
implementations ?

I found two : "SMB for Fuse" and "usmb". I haven't had time to try this
out yet.

Regards,


Fabien


ales...@seznam.cz a écrit :
> Hello,
>
> I can confirm that. I've tried smbclient from Samba 3.3.0rc2. It is
considerably faster than cifs kernel module (version 1.52 from Linux 2.6.25).
The file system is exported via Samba 3.0.33. Reading 700MB file (residing in
the buffer cache) from the server I get:
>
> cifs.ko ~ 30MB/s
> smbclient ~ 74MB/s
>
> I needed to remove the NIC IRQ affinity to only one CPU as it got completely
saturated at about 66MB/s (1GHz Pentium III). Still think that the machine would
do more than 74MB/s, but it is much better than 30MB/s anyway. Unfortunately I
need the share mounted as a regular file system, so I'm gonna stick with the
kernel implementation. One day I will try using jumbo frames, but for the moment
I'm stuck with 1500b frames. In regards to the question Fabien has raised - I
think Steven French has the performance optimization in TODO section for the
cifs module, but he only knows when he gets there. Haven't asked him though.
Anyway thanks for the help.
>
> Regards
>
> Aleš
>
> P.S.: The performance problems of various SMB/CIFS clients should probably be
mentioned in Samba docs, that would prevent people from asking the same
questions over and over again. Might even force Steven to fix the cifs kernel
module ;-)
>
>> ---- Původní zpráva 
>> Od: Volker Lendecke 
>> Předmět: Re: [Samba] Samba performance issue
>> Datum: 05.1.2009 21:42:42
>> 
>> On Mon, Jan 05, 2009 at 08:25:34PM +0100, Fabien wrote:
>>> I've seen I'm not the only one impacted with this issue these times on
>>> the mailing list :)
>>>
>>> I did the following test (Debian packages) :
>>>
>>> Server & Client : samba 3.2.5
>>> mount -t smbfs : ~35Mo/s
>>> mount -t cifs : ~35Mo/s
>>> smbclient : ~80Mo/s
>>>
>>> Server & Client : samba 3.0.24
>>> mount -t smbfs : ~35Mo/s
>>> mount -t cifs : ~35Mo/s
>>> smbclient : ~60Mo/s
>>>
>>> This is the first time I try smbclient.
>>>
>>> There is a real big difference between mount and smbclient !
>>>
>>> And it seems to be better to use the 3.2.5 version which is ~ 20Mo/s
>>> better than the 3.0.24 version.
>>>
>>> Again, all of this was tested without using the disks (buffer cache).
>>>
>>> Do you know where does this difference comes from ?
>> It's the latencies that kill performance. Given the
>> request->response nature of the protocol with a limited
>> request size (no matter how large you make them), you can
>> only get a certain number of round trips per second.
>> smbclient 3.2 and even more in upcoming 3.3 hides those
>> latencies by issuing more than one request at the same time
>> using the "Multiplex ID" field in the SMB header properly.
>> Neither cifs nor smbfs do this.
>>
>> Volker
>>
>>
>>
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba




--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Fabien]

Hello,

as you say, I also think it would be nice to mention the issue in the 
documentation to avoid people turning crazy !


If you need the filesystem to be mounted, did you try the fuse 
implementations ?


I found two : "SMB for Fuse" and "usmb". I haven't had time to try this 
out yet.


Regards,


Fabien


ales...@seznam.cz a écrit :

Hello,

I can confirm that. I've tried smbclient from Samba 3.3.0rc2. It is 
considerably faster than cifs kernel module (version 1.52 from Linux 2.6.25). 
The file system is exported via Samba 3.0.33. Reading 700MB file (residing in 
the buffer cache) from the server I get:

cifs.ko ~ 30MB/s
smbclient ~ 74MB/s

I needed to remove the NIC IRQ affinity to only one CPU as it got completely 
saturated at about 66MB/s (1GHz Pentium III). Still think that the machine 
would do more than 74MB/s, but it is much better than 30MB/s anyway. 
Unfortunately I need the share mounted as a regular file system, so I'm gonna 
stick with the kernel implementation. One day I will try using jumbo frames, 
but for the moment I'm stuck with 1500b frames. In regards to the question 
Fabien has raised - I think Steven French has the performance optimization in 
TODO section for the cifs module, but he only knows when he gets there. Haven't 
asked him though. Anyway thanks for the help.

Regards

Aleš 


P.S.: The performance problems of various SMB/CIFS clients should probably be 
mentioned in Samba docs, that would prevent people from asking the same 
questions over and over again. Might even force Steven to fix the cifs kernel 
module ;-)


 Původní zpráva --------
Od: Volker Lendecke 
Předmět: Re: [Samba] Samba performance issue
Datum: 05.1.2009 21:42:42

On Mon, Jan 05, 2009 at 08:25:34PM +0100, Fabien wrote:
I've seen I'm not the only one impacted with this issue these times on 
the mailing list :)


I did the following test (Debian packages) :

Server & Client : samba 3.2.5
mount -t smbfs : ~35Mo/s
mount -t cifs : ~35Mo/s
smbclient : ~80Mo/s

Server & Client : samba 3.0.24
mount -t smbfs : ~35Mo/s
mount -t cifs : ~35Mo/s
smbclient : ~60Mo/s

This is the first time I try smbclient.

There is a real big difference between mount and smbclient !

And it seems to be better to use the 3.2.5 version which is ~ 20Mo/s 
better than the 3.0.24 version.


Again, all of this was tested without using the disks (buffer cache).

Do you know where does this difference comes from ?

It's the latencies that kill performance. Given the
request->response nature of the protocol with a limited
request size (no matter how large you make them), you can
only get a certain number of round trips per second.
smbclient 3.2 and even more in upcoming 3.3 hides those
latencies by issuing more than one request at the same time
using the "Multiplex ID" field in the SMB header properly.
Neither cifs nor smbfs do this.

Volker




--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[ales-76]

Hello,

I can confirm that. I've tried smbclient from Samba 3.3.0rc2. It is 
considerably faster than cifs kernel module (version 1.52 from Linux 2.6.25). 
The file system is exported via Samba 3.0.33. Reading 700MB file (residing in 
the buffer cache) from the server I get:

cifs.ko ~ 30MB/s
smbclient ~ 74MB/s

I needed to remove the NIC IRQ affinity to only one CPU as it got completely 
saturated at about 66MB/s (1GHz Pentium III). Still think that the machine 
would do more than 74MB/s, but it is much better than 30MB/s anyway. 
Unfortunately I need the share mounted as a regular file system, so I'm gonna 
stick with the kernel implementation. One day I will try using jumbo frames, 
but for the moment I'm stuck with 1500b frames. In regards to the question 
Fabien has raised - I think Steven French has the performance optimization in 
TODO section for the cifs module, but he only knows when he gets there. Haven't 
asked him though. Anyway thanks for the help.

Regards

Aleš

P.S.: The performance problems of various SMB/CIFS clients should probably be 
mentioned in Samba docs, that would prevent people from asking the same 
questions over and over again. Might even force Steven to fix the cifs kernel 
module ;-)

>  Původní zpráva 
> Od: Volker Lendecke 
> Předmět: Re: [Samba] Samba performance issue
> Datum: 05.1.2009 21:42:42
> 
> On Mon, Jan 05, 2009 at 08:25:34PM +0100, Fabien wrote:
> > I've seen I'm not the only one impacted with this issue these times on
> > the mailing list :)
> >
> > I did the following test (Debian packages) :
> >
> > Server & Client : samba 3.2.5
> > mount -t smbfs : ~35Mo/s
> > mount -t cifs : ~35Mo/s
> > smbclient : ~80Mo/s
> >
> > Server & Client : samba 3.0.24
> > mount -t smbfs : ~35Mo/s
> > mount -t cifs : ~35Mo/s
> > smbclient : ~60Mo/s
> >
> > This is the first time I try smbclient.
> >
> > There is a real big difference between mount and smbclient !
> >
> > And it seems to be better to use the 3.2.5 version which is ~ 20Mo/s
> > better than the 3.0.24 version.
> >
> > Again, all of this was tested without using the disks (buffer cache).
> >
> > Do you know where does this difference comes from ?
>
> It's the latencies that kill performance. Given the
> request->response nature of the protocol with a limited
> request size (no matter how large you make them), you can
> only get a certain number of round trips per second.
> smbclient 3.2 and even more in upcoming 3.3 hides those
> latencies by issuing more than one request at the same time
> using the "Multiplex ID" field in the SMB header properly.
> Neither cifs nor smbfs do this.
>
> Volker
>
>
>
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Jeremy Allison]

On Thu, Jan 08, 2009 at 09:24:16AM -0800, rhubbell wrote:
> On Thu, 2009-01-08 at 08:27 -0800, Jeremy Allison wrote:
> > On Thu, Jan 08, 2009 at 03:27:53PM +0100, Fabien wrote:
> > > Thanks for the information.
> > >
> > > Do you know why the smbclient, although faster, is not fast enough to go  
> > > over 80Mo/s ?
> > >
> > > Is there any plan to do the fiddly work on the smbfs implementation to  
> > > make it as fast as smbclient ? :)
> > 
> > smbfs is dead. CIFSFS is under active development.
> 
> 
> Is the fiddly work being done in CIFSFS?  Or planned?

Talk to Steve French and Jeff Layton, who are the most
active developers on CIFSFS.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[rhubbell]

On Thu, 2009-01-08 at 08:27 -0800, Jeremy Allison wrote:
> On Thu, Jan 08, 2009 at 03:27:53PM +0100, Fabien wrote:
> > Thanks for the information.
> >
> > Do you know why the smbclient, although faster, is not fast enough to go  
> > over 80Mo/s ?
> >
> > Is there any plan to do the fiddly work on the smbfs implementation to  
> > make it as fast as smbclient ? :)
> 
> smbfs is dead. CIFSFS is under active development.


Is the fiddly work being done in CIFSFS?  Or planned?


-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Jeremy Allison]

On Thu, Jan 08, 2009 at 03:27:53PM +0100, Fabien wrote:
> Thanks for the information.
>
> Do you know why the smbclient, although faster, is not fast enough to go  
> over 80Mo/s ?
>
> Is there any plan to do the fiddly work on the smbfs implementation to  
> make it as fast as smbclient ? :)

smbfs is dead. CIFSFS is under active development.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Thu, Jan 08, 2009 at 03:27:53PM +0100, Fabien wrote:
> Thanks for the information.
> 
> Do you know why the smbclient, although faster, is not fast enough to go 
> over 80Mo/s ?

No, not from the top of my head. This needs much closer
investigation.

Volker


pgpCbC6MT7n10.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Fabien]

Thanks for the information.

Do you know why the smbclient, although faster, is not fast enough to go 
over 80Mo/s ?


Is there any plan to do the fiddly work on the smbfs implementation to 
make it as fast as smbclient ? :)


I didn't try the fuse implemtations yet. I found two : "SMB for Fuse" 
and "usmb". I going to try them when possible.


Has anyone already tried them ?

Thanks !


Fabien

Volker Lendecke a écrit :

I did the following test (Debian packages) :

Server & Client : samba 3.2.5
mount -t smbfs : ~35Mo/s
mount -t cifs : ~35Mo/s
smbclient : ~80Mo/s

Server & Client : samba 3.0.24
mount -t smbfs : ~35Mo/s
mount -t cifs : ~35Mo/s
smbclient : ~60Mo/s


It's the latencies that kill performance. Given the
request->response nature of the protocol with a limited
request size (no matter how large you make them), you can
only get a certain number of round trips per second.
smbclient 3.2 and even more in upcoming 3.3 hides those
latencies by issuing more than one request at the same time
using the "Multiplex ID" field in the SMB header properly.
Neither cifs nor smbfs do this.

Volker

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Helmut Hullen]

Hallo, wes,

Du (samba) meintest am 06.01.09:

> Can I use smbclient to
> create a mount point on the unix filesystem the way I can with
> mount.cifs?

That's the job of "mkdir". No other program.

Viele Gruesse!
Helmut
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Tue, Jan 06, 2009 at 11:45:33AM -0800, wes wrote:
> > Ha, lol, no.  My question was probably ridiculous beyond comprehension.
> > Was asking if there was a way to make use of
> > smbclient to replace cifs or smbfs.
> >
> 
> I would also like to know the answer to this. Can I use smbclient to create
> a mount point on the unix filesystem the way I can with mount.cifs?

No, you can't. Your options are mount.cifs, mount.smbfs,
sharity (nfs2smb converter) and some fuse-based file
systems. And then there are the pure user-space VFS
implementations of gnome and kde and possibly others. But
they only work for the applications using them.

Volker


pgpubbDt8pwHy.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[wes]

>
> > > Is there a way to setup smbclient to act like a mount point acts? I'm
> > > pretty sure the answer's "No." but I ask anyway.
> >
> > What do you mean by that? You want to slow down smbclient?
>
> Ha, lol, no.  My question was probably ridiculous beyond comprehension.
> Was asking if there was a way to make use of
> smbclient to replace cifs or smbfs.
>

I would also like to know the answer to this. Can I use smbclient to create
a mount point on the unix filesystem the way I can with mount.cifs?

thanks,
-wes
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[rhubbell]

On Tue, 2009-01-06 at 20:25 +0100, Volker Lendecke wrote:
> On Tue, Jan 06, 2009 at 10:41:55AM -0800, rhubbell wrote:
> > > > Why do cifs and smbfs not have this capability?  Is it too much work? Is
> > > > it due to differences in the purpose of each?
> > > 
> > > It's fiddly work nobody has done yet.
> > 
> > fiddly = not hard work, but tedious and sort of annoying?
> 
> Fiddly as in not many of lines of code, but code with
> complex interactions and data dependencies. So I'd say for
> me it is hard work, others might find it easier. But as
> nobody has done it yet, I'd say I'm not completely alone in
> that assessment.


Ah, ok, understood, thanks. So it falls into the class of enhancements
under the heading "Worthy but tricky". 

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Tue, Jan 06, 2009 at 10:41:55AM -0800, rhubbell wrote:
> > > Why do cifs and smbfs not have this capability?  Is it too much work? Is
> > > it due to differences in the purpose of each?
> > 
> > It's fiddly work nobody has done yet.
> 
> fiddly = not hard work, but tedious and sort of annoying?

Fiddly as in not many of lines of code, but code with
complex interactions and data dependencies. So I'd say for
me it is hard work, others might find it easier. But as
nobody has done it yet, I'd say I'm not completely alone in
that assessment.

Volker


pgpvVWEcp88kz.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[rhubbell]

On Tue, 2009-01-06 at 19:20 +0100, Volker Lendecke wrote:
> On Tue, Jan 06, 2009 at 09:35:39AM -0800, rhubbell wrote:
> > Why do cifs and smbfs not have this capability?  Is it too much work? Is
> > it due to differences in the purpose of each?
> 
> It's fiddly work nobody has done yet.


fiddly = not hard work, but tedious and sort of annoying?

> 
> > Is there a way to setup smbclient to act like a mount point acts? I'm
> > pretty sure the answer's "No." but I ask anyway.
> 
> What do you mean by that? You want to slow down smbclient?

Ha, lol, no.  My question was probably ridiculous beyond comprehension.
Was asking if there was a way to make use of
smbclient to replace cifs or smbfs.

> 
> > Is this definition correct?
> > 
> > Multiplex ID:
> > Used by the server to verify the file access permissions of groups where
> > consumer-based file protection is in effect.
> 
> No, that is not correct. See
> http://go.microsoft.com/fwlink/?LinkId=89836 for a
> description of CIFS, alternatively look at
> http://ubiqx.org/cifs.


Thanks for those links, that definition I found definitely had me
scratching my head.


> 
> Volker

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Tue, Jan 06, 2009 at 09:35:39AM -0800, rhubbell wrote:
> Why do cifs and smbfs not have this capability?  Is it too much work? Is
> it due to differences in the purpose of each?

It's fiddly work nobody has done yet.

> Is there a way to setup smbclient to act like a mount point acts? I'm
> pretty sure the answer's "No." but I ask anyway.

What do you mean by that? You want to slow down smbclient?

> Is this definition correct?
> 
> Multiplex ID:
> Used by the server to verify the file access permissions of groups where
> consumer-based file protection is in effect.

No, that is not correct. See
http://go.microsoft.com/fwlink/?LinkId=89836 for a
description of CIFS, alternatively look at
http://ubiqx.org/cifs.

Volker


pgpnhqJxuFGqb.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[rhubbell]

On Mon, 2009-01-05 at 21:47 +0100, Volker Lendecke wrote:
> On Mon, Jan 05, 2009 at 08:25:34PM +0100, Fabien wrote:
> > I've seen I'm not the only one impacted with this issue these times on 
> > the mailing list :)
> > 
> > I did the following test (Debian packages) :
> > 
> > Server & Client : samba 3.2.5
> > mount -t smbfs : ~35Mo/s
> > mount -t cifs : ~35Mo/s
> > smbclient : ~80Mo/s
> > 
> > Server & Client : samba 3.0.24
> > mount -t smbfs : ~35Mo/s
> > mount -t cifs : ~35Mo/s
> > smbclient : ~60Mo/s
> > 
> > This is the first time I try smbclient.
> > 
> > There is a real big difference between mount and smbclient !
> > 
> > And it seems to be better to use the 3.2.5 version which is ~ 20Mo/s 
> > better than the 3.0.24 version.
> > 
> > Again, all of this was tested without using the disks (buffer cache).
> > 
> > Do you know where does this difference comes from ?
> 
> It's the latencies that kill performance. Given the
> request->response nature of the protocol with a limited
> request size (no matter how large you make them), you can
> only get a certain number of round trips per second.
> smbclient 3.2 and even more in upcoming 3.3 hides those
> latencies by issuing more than one request at the same time
> using the "Multiplex ID" field in the SMB header properly.
> Neither cifs nor smbfs do this.


Why do cifs and smbfs not have this capability?  Is it too much work? Is
it due to differences in the purpose of each?

Is there a way to setup smbclient to act like a mount point acts? I'm
pretty sure the answer's "No." but I ask anyway.

Is this definition correct?

Multiplex ID:
Used by the server to verify the file access permissions of groups where
consumer-based file protection is in effect.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Mon, Jan 05, 2009 at 08:25:34PM +0100, Fabien wrote:
> I've seen I'm not the only one impacted with this issue these times on 
> the mailing list :)
> 
> I did the following test (Debian packages) :
> 
> Server & Client : samba 3.2.5
> mount -t smbfs : ~35Mo/s
> mount -t cifs : ~35Mo/s
> smbclient : ~80Mo/s
> 
> Server & Client : samba 3.0.24
> mount -t smbfs : ~35Mo/s
> mount -t cifs : ~35Mo/s
> smbclient : ~60Mo/s
> 
> This is the first time I try smbclient.
> 
> There is a real big difference between mount and smbclient !
> 
> And it seems to be better to use the 3.2.5 version which is ~ 20Mo/s 
> better than the 3.0.24 version.
> 
> Again, all of this was tested without using the disks (buffer cache).
> 
> Do you know where does this difference comes from ?

It's the latencies that kill performance. Given the
request->response nature of the protocol with a limited
request size (no matter how large you make them), you can
only get a certain number of round trips per second.
smbclient 3.2 and even more in upcoming 3.3 hides those
latencies by issuing more than one request at the same time
using the "Multiplex ID" field in the SMB header properly.
Neither cifs nor smbfs do this.

Volker


pgpPjH7wyRn3x.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Fabien]

Hello,

I've seen I'm not the only one impacted with this issue these times on 
the mailing list :)


I did the following test (Debian packages) :

Server & Client : samba 3.2.5
mount -t smbfs : ~35Mo/s
mount -t cifs : ~35Mo/s
smbclient : ~80Mo/s

Server & Client : samba 3.0.24
mount -t smbfs : ~35Mo/s
mount -t cifs : ~35Mo/s
smbclient : ~60Mo/s

This is the first time I try smbclient.

There is a real big difference between mount and smbclient !

And it seems to be better to use the 3.2.5 version which is ~ 20Mo/s 
better than the 3.0.24 version.


Again, all of this was tested without using the disks (buffer cache).

Do you know where does this difference comes from ?

Fabien


Volker Lendecke a écrit :

On Sat, Jan 03, 2009 at 02:26:01AM +0100, Fabien wrote:

I'm gonna try that and post the results here as soon as possible.

Do you think it could really make a difference knowing that I also tried 
the WindowsXP native client without being able to notice any difference ?


Not sure, but for me it makes a difference.

Volker

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Fabien]

I'm gonna try that and post the results here as soon as possible.

Do you think it could really make a difference knowing that I also tried
the WindowsXP native client without being able to notice any difference ?

I must also say that I used cifs for my tests (mount -t cifs).

Thanks again,


Fabien


Volker Lendecke a écrit :

On Thu, Jan 01, 2009 at 07:35:06PM +0100, Fabien wrote:

* My server config :
   - AMD X2 4200+
   - 2 Go RAM
   - 4 x 500 Go --> RAID5
   - Gigabyte connection
   - Debian ETCH
   - debian package : Samba 3.0.24 (I also tried to backport the testing
version => 3.2.5 but the results were exactly the same)


Can you try the smbclient from 3.2.5 to get a large file and
see if that performs better?

Thanks,

Volker


--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Aleš Bláha]

On Fri, 02 Jan 2009 22:01:49 +0100 (CET)
Volker Lendecke  wrote:

> On Fri, Jan 02, 2009 at 09:57:34PM +0100, Aleš Bláha wrote:
> > Thank you for your help. I will try what you propose as soon as I get
> > to the machines. But, to be honest, I don't think, the hardware is the
> > bottleneck. The RAID controller and the NIC in the server sit on a
> > different PCI bus and each one has its interrupt hooked to a
> > different CPU. Appart from that, as I've mentioned in the previous
> > post, I can saturate the network when copying files that are cached on
> > the server - but only as long as there are several pending requests -
> > with only one I get just those 30MBs or so (one tcp session vs. couple
> > of them). I will do some more benchmarks next week, post the
> > smb.conf and tcp/ip stack config.
> 
> One thing you might also want to check is get/put operations
> with latest smbclient from the git master branch. This does
> almost optimal streaming smb operations.
> 
> Volker
> 

Hello Volker,

I will try that also, that is after I find out how to build it from the git 
source.

Thanks

Ales
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Fri, Jan 02, 2009 at 09:57:34PM +0100, Aleš Bláha wrote:
> Thank you for your help. I will try what you propose as soon as I get
> to the machines. But, to be honest, I don't think, the hardware is the
> bottleneck. The RAID controller and the NIC in the server sit on a
> different PCI bus and each one has its interrupt hooked to a
> different CPU. Appart from that, as I've mentioned in the previous
> post, I can saturate the network when copying files that are cached on
> the server - but only as long as there are several pending requests -
> with only one I get just those 30MBs or so (one tcp session vs. couple
> of them). I will do some more benchmarks next week, post the
> smb.conf and tcp/ip stack config.

One thing you might also want to check is get/put operations
with latest smbclient from the git master branch. This does
almost optimal streaming smb operations.

Volker


pgpFcGD0pFnnJ.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Aleš Bláha]

Hi Volker,

Thank you for your help. I will try what you propose as soon as I get
to the machines. But, to be honest, I don't think, the hardware is the
bottleneck. The RAID controller and the NIC in the server sit on a
different PCI bus and each one has its interrupt hooked to a
different CPU. Appart from that, as I've mentioned in the previous
post, I can saturate the network when copying files that are cached on
the server - but only as long as there are several pending requests -
with only one I get just those 30MBs or so (one tcp session vs. couple
of them). I will do some more benchmarks next week, post the
smb.conf and tcp/ip stack config.

Regards

Ales


On Fri, 02 Jan 2009 20:48:53 +0100 (CET)
Volker Lendecke  wrote:

> On Fri, Jan 02, 2009 at 05:54:11PM +0100, Aleš Bláha wrote:
> > Both computers run Gentoo Linux 2008, kernel 2.6.25-r9,
> > server runs Samba 3.0.33, client mount.cifs 3.0.30.  The
> > underlying filesystem for Samba is Ext3 with xattr and
> > acls. I wasn't able to break 32MB/s (250Mbps) transfer
> > speed neither reading nor writing to the server. The disk
> > subsystem of the server is capable of 60MB/s and generaly
> > the hardware is not the bottleneck. Neither is the network
> > - the bw_tcp from LMbench suite shows around 108MB/s with
> > 1500b messages, which is what I would expect from GbE &
> > TCP/IP. I've been tinkering with
> 
> In a test I did lately it made a huge difference if I just
> did raw TCP benchmarks, raw disk benchmarks or a combined
> one. The test I used was
> 
> netcat -l -p  > diskfile
> 
> on the receiving end and
> 
> netcat    
> on the sending end. This made my hardware which would
> otherwise happily saturate gigE crawl down to something like
> 50MB/sec. Can you try that?
> 
> Volker
> 


-- 

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Fri, Jan 02, 2009 at 05:54:11PM +0100, Aleš Bláha wrote:
> Both computers run Gentoo Linux 2008, kernel 2.6.25-r9,
> server runs Samba 3.0.33, client mount.cifs 3.0.30.  The
> underlying filesystem for Samba is Ext3 with xattr and
> acls. I wasn't able to break 32MB/s (250Mbps) transfer
> speed neither reading nor writing to the server. The disk
> subsystem of the server is capable of 60MB/s and generaly
> the hardware is not the bottleneck. Neither is the network
> - the bw_tcp from LMbench suite shows around 108MB/s with
> 1500b messages, which is what I would expect from GbE &
> TCP/IP. I've been tinkering with

In a test I did lately it made a huge difference if I just
did raw TCP benchmarks, raw disk benchmarks or a combined
one. The test I used was

netcat -l -p  > diskfile

on the receiving end and

netcat   

pgpjKUD24NYHS.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[Volker Lendecke]

On Thu, Jan 01, 2009 at 07:35:06PM +0100, Fabien wrote:
> * My server config :
>- AMD X2 4200+
>- 2 Go RAM
>- 4 x 500 Go --> RAID5
>- Gigabyte connection
>- Debian ETCH
>- debian package : Samba 3.0.24 (I also tried to backport the testing
> version => 3.2.5 but the results were exactly the same)

Can you try the smbclient from 3.2.5 to get a large file and
see if that performs better?

Thanks,

Volker


pgptHifpmfoVo.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance degrade

[vishesh kumar]

Thanks for attention

 But there is no networking related issue, the server is perfectly
accessible for ftp server. Also there is no RTO for ping to samba server.
The only new thing that i done is configured DNS slave server on that.

On Wed, Oct 22, 2008 at 10:28 PM, Hoover, Tony <[EMAIL PROTECTED]> wrote:

> have you ruled out a networking problem? (i.e. Switch didn't auto-neg to
> the
> same speed/duplex settings as the server)?
>
>
>  
> Tony Hoover, Network Administrator
> KSU - Salina, College of Technology and Aviation
> (785) 826-2660
>
> "Don't Blend in..."
> 
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:samba-bounces+hoover =sal.ksu.edu@
> lists.samba.org] On Behalf Of
> vishesh
> Sent: Wednesday, October 22, 2008 4:55 PM
> To: samba@lists.samba.org
> Subject: [Samba] samba performance degrade
>
> dear all
>  I am using samba 3.0.28 on RHEl 5.2. I am using samba with winbind that
> authenticate window 2003 domain. From last few months samba was running
> properly. But today samba server performance badly, even sometime mapped
> drive on xp disappear. When i tried to connect samba shares, error appear
> server not available. Anyone suggest me what may be the problem. Why samba
> performance degrade after running around a month.
>
> my configuration is as follows
>
> #=== Global Settings
> =
>
> [global]
> #--authconfig--start-line--
>
> # Generated by authconfig on 2008/09/04 22:25:21 # DO NOT EDIT THIS SECTION
> (delimited by --start-line--/--end-line--) # Any modification may be
> deleted
> or altered by authconfig in future
>
>   workgroup = abp
>   password server = s2.abp.del
>   realm = ABP.DEL
>   security = ads
>   idmap uid = 16777216-33554431
>   idmap gid = 16777216-33554431
>   template shell = /bin/bash
>   winbind use default domain = true
>   winbind offline logon = false
>   winbind enum users = yes
>   winbind enum groups = yes
>
> #--authconfig--end-line--
> server string =
> netbios name = abpdel2
> admin users = @"abp\domain admins"
> # --- Network Related Options -
> #
>
>
>passdb backend = tdbsam
>
>
>
>  map read only = no
> ;map system = no
> ;store dos attributes = yes
> hide dot files = yes
> veto files=/lost+found/Trash/Recycler
> # Share Definitions
> ==
>
> ;[homes]
> ;comment = Home Directories
> ;browseable = no
> ;writeable = yes
> ;valid users = %S
> ;valid users = MYDOMAIN\%S
>
> ;[printers]
> ;comment = All Printers
> ;path = /var/spool/samba
> ;browseable = no
> ;guest ok = no
> ;writeable = no
> ;printable = yes
>
> # Un-comment the following and create the netlogon directory for Domain
> Logons
> ;[netlogon]
> ;comment = Network Logon Service
> ;path = /var/lib/samba/netlogon
> ;guest ok = yes
> ;writable = no
> ;share modes = no
>
>
> # Un-comment the following to provide a specific roving profile share
> # the default is to use the user's home directory
> ;[Profiles]
> ;path = /var/lib/samba/profiles
> ;browseable = no
> ;guest ok = yes
>
>
> # A publicly accessible directory, but read only, except for people in
> # the "staff" group
> ;[public]
> ;comment = Public Stuff
> ;path = /home/samba
> ;public = yes
> ;writable = yes
> ;printable = no
> ;write list = +staff
>
> [design]
>comment = "home folder for design department"
>path = /data/design
>vfs object = recycle
>recycle:repository = /data/design/Trash
>recycle:keeptrace = Yes
>writeable = yes
>create mask = 644
>directory mask = 2755
> [home]
>comment = "home folder for bwedit"
>path = /data/edit/home
>vfs object = recycle
>recycle:repository = /data/edit/home/Trash
>recycle:keeptrace = Yes
>writeable = yes
>create mask = 644
>directory mask = 2755
>
> [user]
>comment = " home directory to tt feature and ab edit users"
>path = /data/edit/user
>vfs object = recycle
>recycle:repository = /data/edit/user/Trash
>recycle:keeptrace = Yes
>writeable = yes
>create mask = 644
>directory mask = 2755
> [mark]
>comment = "marketting users"
>path = /data/marketting
>vfs object = recycle
>vfs:repository = /data/marketting/Trash
>vfs:keeptrace = Yes
>writeable = yes
>create mask = 644
>directory mask = 2755
>
> [system]
>comment = "home folder for it"
>path = /data/system
>vfs object = recycle
>recycle:repository = /data/system/Trash
>recycle:keeptrace = Yes
>writeable = yes
>
> [com]
>path=/data/marketting/com
>vfs object = recycle
>recycle:repository = /data/marketting/Trash
>recycle:keeptrace = Yes
>writeabl

RE: [Samba] samba performance degrade

[Hoover, Tony]

have you ruled out a networking problem? (i.e. Switch didn't auto-neg to the
same speed/duplex settings as the server)? 


 
Tony Hoover, Network Administrator
KSU - Salina, College of Technology and Aviation
(785) 826-2660

"Don't Blend in..."


-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of
vishesh
Sent: Wednesday, October 22, 2008 4:55 PM
To: samba@lists.samba.org
Subject: [Samba] samba performance degrade

dear all
  I am using samba 3.0.28 on RHEl 5.2. I am using samba with winbind that
authenticate window 2003 domain. From last few months samba was running
properly. But today samba server performance badly, even sometime mapped
drive on xp disappear. When i tried to connect samba shares, error appear
server not available. Anyone suggest me what may be the problem. Why samba
performance degrade after running around a month.

my configuration is as follows

#=== Global Settings
=

[global]
#--authconfig--start-line--

# Generated by authconfig on 2008/09/04 22:25:21 # DO NOT EDIT THIS SECTION
(delimited by --start-line--/--end-line--) # Any modification may be deleted
or altered by authconfig in future

   workgroup = abp
   password server = s2.abp.del
   realm = ABP.DEL
   security = ads
   idmap uid = 16777216-33554431
   idmap gid = 16777216-33554431
   template shell = /bin/bash
   winbind use default domain = true
   winbind offline logon = false
   winbind enum users = yes
   winbind enum groups = yes

#--authconfig--end-line--
server string =
netbios name = abpdel2
admin users = @"abp\domain admins"
# --- Network Related Options -
#


passdb backend = tdbsam



 map read only = no
;map system = no
;store dos attributes = yes
hide dot files = yes
veto files=/lost+found/Trash/Recycler
# Share Definitions
==

;[homes]
;comment = Home Directories
;browseable = no
;writeable = yes
;valid users = %S
;valid users = MYDOMAIN\%S

;[printers]
;comment = All Printers
;path = /var/spool/samba
;browseable = no
;guest ok = no
;writeable = no
;printable = yes

# Un-comment the following and create the netlogon directory for Domain
Logons
;[netlogon]
;comment = Network Logon Service
;path = /var/lib/samba/netlogon
;guest ok = yes
;writable = no
;share modes = no


# Un-comment the following to provide a specific roving profile share
# the default is to use the user's home directory
;[Profiles]
;path = /var/lib/samba/profiles
;browseable = no
;guest ok = yes


# A publicly accessible directory, but read only, except for people in
# the "staff" group
;[public]
;comment = Public Stuff
;path = /home/samba
;public = yes
;writable = yes
;printable = no
;write list = +staff

[design]
comment = "home folder for design department"
path = /data/design
vfs object = recycle
recycle:repository = /data/design/Trash
recycle:keeptrace = Yes
writeable = yes
create mask = 644
directory mask = 2755
[home]
comment = "home folder for bwedit"
path = /data/edit/home
vfs object = recycle
recycle:repository = /data/edit/home/Trash
recycle:keeptrace = Yes
writeable = yes
create mask = 644
directory mask = 2755

[user]
comment = " home directory to tt feature and ab edit users"
path = /data/edit/user
vfs object = recycle
recycle:repository = /data/edit/user/Trash
recycle:keeptrace = Yes
writeable = yes
create mask = 644
directory mask = 2755
[mark]
comment = "marketting users"
path = /data/marketting
vfs object = recycle
vfs:repository = /data/marketting/Trash
vfs:keeptrace = Yes
writeable = yes
create mask = 644
directory mask = 2755

[system]
comment = "home folder for it"
path = /data/system
vfs object = recycle
recycle:repository = /data/system/Trash
recycle:keeptrace = Yes
writeable = yes

[com]
path=/data/marketting/com
vfs object = recycle
recycle:repository = /data/marketting/Trash
recycle:keeptrace = Yes
writeable=yes

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance tuning

[John Drescher]

On 10/9/07, Daniel L. Miller <[EMAIL PROTECTED]> wrote:
> Andrew Sherlock-CF wrote:
> > I wonder if tshark or netstat could be useful here
> I don't know the tools, which is why I was asking.
>
I do not think either tool will benchmark samba file serving
performance but the actual goal of what a  samba mark would benchmark
is unclear. I guess one could create a benchmark that could measure
the throughput of a single folder on a single samba share but this may
or may not be a good guide to performance on the samba server. I mean
permissions, number of files in a folder, file sizes, acls, and file
system and storage choices can play a big part in performance and any
or all of these may be different from folder to folder on a samba
share.

John
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance tuning

[Daniel L. Miller]

Andrew Sherlock-CF wrote:

I wonder if tshark or netstat could be useful here

I don't know the tools, which is why I was asking.

Daniel
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba performance tuning

[Andrew Sherlock-CF]

I wonder if tshark or netstat could be useful here?

Andrew 
--- 

> -Original Message-
> From: [EMAIL PROTECTED] 
> [mailto:[EMAIL PROTECTED]
> g] On Behalf Of Daniel L. Miller
> Sent: 09 October 2007 00:47
> To: samba@lists.samba.org
> Subject: [Samba] Samba performance tuning
> 
> Hi all!
> 
> I've seen a number of posts regarding Samba performance - either 
> comments about perceived poor performance, or recommended parameter 
> changes.  Instead of some arbitrary buffer numbers, or "play with it 
> until it works", are there any analysis tools that can give 
> quantitative 
> answers ?  Something where instead of just saying, "Samba is 
> too slow", 
> I could say, "on a quad-opteron system, with 4G RAM, and full-duplex 
> 100BaseT network, SambaMark reports a score of 237.".
> 
> Maybe that's too much of a dream - is there a particular file 
> size/copy/read technique and associated analysis that I should try?
> -- 
> Daniel
> -- 
> To unsubscribe from this list go to the following URL and read the
> instructions:  https://lists.samba.org/mailman/listinfo/samba
> 

http://www.bbc.co.uk/
This e-mail (and any attachments) is confidential and may contain personal 
views which are not the views of the BBC unless specifically stated.
If you have received it in error, please delete it from your system.
Do not use, copy or disclose the information in any way nor act in reliance on 
it and notify the sender immediately.
Please note that the BBC monitors e-mails sent or received.
Further communication will signify your consent to this.

--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance metrics

[Scott Feldstein]

I have found this, which looks like it could be interesting ->
http://samba.org/samba/docs/man/Samba-HOWTO-Collection/install.html  
(search for perfmon in the page)


It seems like this dir should have perf counters, but I don't see  
them in my running instance.  Anyone know about this?


On my instance it is:

/var/cache/samba/perfmon/*.tbd

But I don't see anything.

Any idea?

On Aug 1, 2007, at 7:28 AM, Dag Wieers wrote:


On Mon, 30 Jul 2007, Scott Feldstein wrote:

I am attempting to write a monitor to gather Samba Server  
performance metrics.
I would like to get raw metrics directly from the server itself  
rather than a
3rd party tool.  Could anyone point me to docs or enlighten me on  
how to

accomplish this?

I am looking to determine the availability of the server (up or  
down), the
number of files requested, any caching information available and  
anything else
which would be relevant to determining the overall health of the  
server.


Only way I know of is smbstatus. If there was an interface to  
smbstatus
where one could pipe commands and you get the results I could use  
that as

a workable interface to connect "dstat" to a local samba server.

See:
http://dag.wieers.com/home-made/dstat/

Sadly, forking smbstatus and grepping output for a few counters is  
more

work than it should be.

So I'm interested as well to learn how to get counters from samba,  
other

than smbstatus that is :)

Kind regards,
--   dag wieers,  [EMAIL PROTECTED],  http://dag.wieers.com/   --
[Any errors in spelling, tact or fact are transmission errors]


--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance metrics

[Dag Wieers]

On Mon, 30 Jul 2007, Scott Feldstein wrote:

> I am attempting to write a monitor to gather Samba Server performance metrics.
> I would like to get raw metrics directly from the server itself rather than a
> 3rd party tool.  Could anyone point me to docs or enlighten me on how to
> accomplish this?
> 
> I am looking to determine the availability of the server (up or down), the
> number of files requested, any caching information available and anything else
> which would be relevant to determining the overall health of the server.

Only way I know of is smbstatus. If there was an interface to smbstatus 
where one could pipe commands and you get the results I could use that as 
a workable interface to connect "dstat" to a local samba server.

See:
http://dag.wieers.com/home-made/dstat/

Sadly, forking smbstatus and grepping output for a few counters is more 
work than it should be.

So I'm interested as well to learn how to get counters from samba, other 
than smbstatus that is :)

Kind regards,
--   dag wieers,  [EMAIL PROTECTED],  http://dag.wieers.com/   --
[Any errors in spelling, tact or fact are transmission errors]
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Volker Lendecke]

On Tue, Feb 20, 2007 at 02:40:41PM +0100, Andrea Lorenz wrote:
> The use of cifsfs increase the read performance from a samba share to a 
> local disk to 30MByte/sec . But the write performace is still 16MByte/sec.
> Do you have any other ideas?

None except to talk to Steve French :-)

Volker


pgpmYTdt8Y8Ms.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Andrea Lorenz]

Hello Volker,

The use of cifsfs increase the read performance from a samba share to a 
local disk to 30MByte/sec . But the write performace is still 16MByte/sec.

Do you have any other ideas?

Andrea


On Fri, Feb 16, 2007 at 02:00:21PM +0100, Andrea Lorenz wrote:

Did you try cifsfs? smbfs is deprecated and orphaned.

If I use mount.cifs I get an error because we use ntlmv2.
We use SC Linux 4.4 with kernel 2.6.9-42.0.3.ELsmp.

Andrea

mount -t cifs -o username=ae106lo,sec=ntlmv2 //guestc1/export1 /mnt_cifs
Password:
mount error 13 = Permission denied


Hmmm. That's a different problem then. No way that for a
speed test you can also allow other auth schemes or even
create a test share that would allow anon access?

The Permission Denied should be solvable though.

Volker


--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Volker Lendecke]

On Fri, Feb 16, 2007 at 02:00:21PM +0100, Andrea Lorenz wrote:
> >Did you try cifsfs? smbfs is deprecated and orphaned.
> 
> If I use mount.cifs I get an error because we use ntlmv2.
> We use SC Linux 4.4 with kernel 2.6.9-42.0.3.ELsmp.
> 
> Andrea
> 
> mount -t cifs -o username=ae106lo,sec=ntlmv2 //guestc1/export1 /mnt_cifs
> Password:
> mount error 13 = Permission denied

Hmmm. That's a different problem then. No way that for a
speed test you can also allow other auth schemes or even
create a test share that would allow anon access?

The Permission Denied should be solvable though.

Volker


pgpzfxVksa8x9.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Andrea Lorenz]

Did you try cifsfs? smbfs is deprecated and orphaned.


If I use mount.cifs I get an error because we use ntlmv2.
We use SC Linux 4.4 with kernel 2.6.9-42.0.3.ELsmp.

Andrea

mount -t cifs -o username=ae106lo,sec=ntlmv2 //guestc1/export1 /mnt_cifs
Password:
mount error 13 = Permission denied


--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Andrea Lorenz]

Did you restart the samba service after changing the socket options?
Yes, of course I have restarted the samba server after changes in  the 
configuration.



Performance will also be dependent on the hardware configuration of
the clients/servers. 
It has been pointed out that with the latest

Linux 2.6 kernels, you need not specify SO_SNDBUF and SO_RCVBUF. Were
you measuring the READ or WRITE performance with this Samba server?
I have testet the server with and without socket options. But it makes 
no significant difference.


Andrea



--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Volker Lendecke]

On Fri, Feb 16, 2007 at 09:47:17AM +0100, Andrea Lorenz wrote:
> If I use CIFS between the linux client(smbmount) or windows client and 
> the samba server I get only 16MByte/sec.

Did you try cifsfs? smbfs is deprecated and orphaned.

Volker


pgp7hFADf7NwI.pgp
Description: PGP signature
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] samba performance

[Srini]

Did you restart the samba service after changing the socket options?
Performance will also be dependent on the hardware configuration of
the clients/servers. It has been pointed out that with the latest
Linux 2.6 kernels, you need not specify SO_SNDBUF and SO_RCVBUF. Were
you measuring the READ or WRITE performance with this Samba server?

srini

On 2/16/07, Andrea Lorenz <[EMAIL PROTECTED]> wrote:

Hello,

I try to find out how can I improve the performance of our samba
installation. For this I install a testbed with a linux server (samba
and nfs), a linux client, a windows 2003 server and a windows XP client.
All computers are connected with gigabit ethernet.

If I use NFS between the linux client and server I get 65MByte/sec,
which is quiet a nice performance.
If I use CIFS between the linux client(smbmount) or windows client and
the samba server I get only 16MByte/sec.
If I use CIFs between the windows client and windows server I get
35MByte/sec. This the more than double speed of the samba server.

What can be the reason for these different speeds? How can I improve the
performace of my samba server. I play around with socket options and max
xmit option. But this does not help. At the moment this values are set:
socket options = TCP_NODELAY IPTOS_LOWDELAY SO_KEEPALIVE
SO_SNDBUF=65536 SO_RCVBUF=65536
max xmit = 65536

I find out, that I can speedup the connection between linux client and
server if I use smbclient with the option -b 15000 . Than I get 30MByte/sec.
How can I set this option with smbmount and on the windows client?

Thanks a lot,
Andrea



--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba


--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance with large directories

[Tomas Baublys]

Hello,

thank you for your help and here is a small summary:
It turned out, that customer application does a wildcard search  " wcard =
c55crvtu.m1>". In windows the kernel does directory filtering. In POSIX,
any wildcard filtering is done in userspace - which means as soon as a
wildcard is received, we have to do a complete directory traversel. If the
client asks for a file using file name samba does not do complete directory
search (since 3.0.12 and using the documented parameters in smb.conf, see
http://us3.samba.org/samba/docs/man/Samba-HOWTO-Collection/largefile.html).
Unfortunately, in this case it asks for multiple files using wildcard and
samba needs to do a userspace match with the wildcard given against the
list of returned names.
A workaround can be written using VFS with new opendir() call, see
http://lists.samba.org/archive/samba-technical/2005-June/041562.html .
We will try to work with customer to change application first.

Tomas Baublys

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance with large directories

[Jeremy Allison]

On Fri, Jun 24, 2005 at 09:04:57PM +0200, Tomas Baublys wrote:
> Jeremy Allison <[EMAIL PROTECTED]> wrote on 24.06.2005 20:32:19:
> 
> > This may be a filesystem problem - looks like you've got the parameters
> > correct. Any chance you can test this on a different filesystem than
> > reiser ? Maybe XFS ?
> >
> > Jeremy.
> 
> I tried XFS, the results are exactly the same. I see following in the logs,
> which looks like the directory is searched file by file:

Please work with me directly on this - I sent you a message off-list.

Thanks,

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance with large directories

[Tomas Baublys]

Jeremy Allison <[EMAIL PROTECTED]> wrote on 24.06.2005 20:32:19:

> This may be a filesystem problem - looks like you've got the parameters
> correct. Any chance you can test this on a different filesystem than
> reiser ? Maybe XFS ?
>
> Jeremy.

I tried XFS, the results are exactly the same. I see following in the logs,
which looks like the directory is searched file by file:

 switch message SMBtrans2 (pid 18061) conn 0x1034c2c0
[2005/06/24 16:17:45, 4] smbd/uid.c:change_to_user(217)
  change_to_user: Skipping user change - already user
[2005/06/24 16:17:45, 3] smbd/trans2.c:call_trans2findfirst(1351)
  call_trans2findfirst: dirtype = 22, maxentries = 170,
close_after_first=0, close_if_end = 0 requires_resume_key = 4 level =
0x104, max_data_bytes = 16384
[2005/06/24 16:17:45, 5] smbd/filename.c:unix_convert(108)
  unix_convert called on file "audiodaten/zentral/c5/c55crvtu.m1>"
[2005/06/24 16:17:45, 5] smbd/filename.c:unix_convert(175)
  unix_convert begin: name = audiodaten/zentral/c5/c55crvtu.m1>, dirpath =
, start = audiodaten/zentral/c5/c55crvtu.m1>
[2005/06/24 16:17:45, 10] smbd/mangle_hash2.c:is_mangled(276)
  is_mangled audiodaten/zentral/c5/c55crvtu.m1> ?
[2005/06/24 16:17:45, 10] smbd/mangle_hash2.c:is_mangled_component(215)
  is_mangled_component audiodaten/zentral/c5/c55crvtu.m1> (len 10) ?
[2005/06/24 16:17:45, 10] smbd/mangle_hash2.c:is_mangled_component(215)
  is_mangled_component audiodaten/zentral/c5/c55crvtu.m1> (len 10) ?
[2005/06/24 16:17:45, 10] smbd/mangle_hash2.c:is_mangled_component(215)
  is_mangled_component zentral/c5/c55crvtu.m1> (len 7) ?
[2005/06/24 16:17:45, 10] smbd/mangle_hash2.c:is_mangled_component(215)
  is_mangled_component c5/c55crvtu.m1> (len 2) ?
[2005/06/24 16:17:45, 10] smbd/mangle_hash2.c:is_mangled_component(215)
  is_mangled_component c55crvtu.m1> (len 12) ?
[2005/06/24 16:17:45, 5] smbd/trans2.c:call_trans2findfirst(1404)
  dir=audiodaten/zentral/c5, mask = c55crvtu.m1>
[2005/06/24 16:17:45, 5] smbd/dir.c:dptr_create(371)
  dptr_create dir=audiodaten/zentral/c5
[2005/06/24 16:17:45, 3] smbd/dir.c:dptr_create(476)
  creating new dirptr 256 for path audiodaten/zentral/c5, expect_close = 1
[2005/06/24 16:17:45, 4] smbd/trans2.c:call_trans2findfirst(1431)
  dptr_num is 256, wcard = c55crvtu.m1>, attr = 22
[2005/06/24 16:17:45, 8] smbd/trans2.c:call_trans2findfirst(1436)
  dirpath= dontdescend=<>
 11320 [2005/06/24 16:17:45, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
  11321   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
4
  11322 [2005/06/24 16:17:45, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
  11323   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
6
  11324 [2005/06/24 16:17:45, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
  11325   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
9
  11326 [2005/06/24 16:17:45, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
  11327   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
12
  11328 [2005/06/24 16:17:45, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
  11329   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
15

=> here it starts and goes like this for 7 seconds until:


 539011   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
794387
 539012 [2005/06/24 16:17:52, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
 539013   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
2147483647
 539014 [2005/06/24 16:17:52, 8] smbd/trans2.c:get_lanman2_dir_entry(869)
 539015   get_lanman2_dir_entry:readdir on dirptr 0x1034b378 now at offset
2147483647
 539016 [2005/06/24 16:17:52, 9] smbd/trans2.c:send_trans2_replies(556)
 539017   t2_rep: params_sent_thistime = 10, data_sent_thistime = 2040,
useable_space = 16582
 539018 [2005/06/24 16:17:52, 9] smbd/trans2.c:send_trans2_replies(558)
 539019   t2_rep: params_to_send = 10, data_to_send = 2040, paramsize = 10,
datasize = 2040
 539020 [2005/06/24 16:17:52, 6] lib/util_sock.c:write_socket(449)
 539021   write_socket(25,2112)
 539022 [2005/06/24 16:17:52, 6] lib/util_sock.c:write_socket(452)
 539023   write_socket(25,2112) wrote 2112
 539024 [2005/06/24 16:17:52, 4] smbd/trans2.c:call_trans2findfirst(1518)
 539025   SMBtrans2 mask=c55crvtu.m1> directory=audiodaten/zentral/c5
dirtype=22 numentries=17
 539026 [2005/06/24 16:17:52, 10]
lib/util_sock.c:read_smb_length_return_keepalive(505)
 539027   got smb length of 37
 539028 [2005/06/24 16:17:52, 6] smbd/process.c:process_smb(1090)
 539029   got message type 0x0 of len 0x25

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance with large directories

[Jeremy Allison]

On Fri, Jun 24, 2005 at 10:42:03AM +0200, Tomas Baublys wrote:
> Hello,
> 
> I have a samba server on OpenPower (SUSE SLES9) and a DS4300 storage
> attached.There  is a lot of data (8 TB mp3 files). Some directories
> contains more then 250.000 files. The application is a music playbox in a
> huge CD shop: you can take a CD hold it under the scanner and after the
> barcode is scanned it starts playing the first CD track (for 30 seconds).
> All files are exactly 125K in size.
> 
> The overall samba performance is very good, but if the source directory
> contains more then 150.000 files, performances decreases. Windows
> application on the client side asks for all file matching the pattern
> 1234ABCD.m* the result is 6-20 Tracks per CD. Using old  NT-Server the
> performance is constant regardless the source directory size.
> Music starts playing after 1 second. Using samba it works good for
> directories below 150.000 files (1s) and it takes 4-6 seconds if the
> directory contains more then 200.000 files to start playing the first
> track. Playing next track starts immediately.
> 
> After some research, I found this is exactly the behaviour corrected by
> Jeremy Allison in February and included in version 3.0.12. I upgraded to
> 3.0.14a, but unfortunately the performance did not improved.
> I am using reiserfs on the data volume and ext3 for /var.
> 
> According to instructions below I am using following options for data
> share:
> 
> case sensitive = True
> default case = lower
> preserve case = no
> short preserve case = no
> 
> http://lists.samba.org/archive/samba/2005-February/099676.html
> http://lists.samba.org/archive/samba/2005-February/099744.html
> http://us3.samba.org/samba/docs/man/Samba-HOWTO-Collection/largefile.html
> 
> 
> I would appreciate any hints which may help to resolve this problem...

This may be a filesystem problem - looks like you've got the parameters
correct. Any chance you can test this on a different filesystem than
reiser ? Maybe XFS ?

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance

[Jeremy Allison]

On Thu, Apr 21, 2005 at 07:25:39PM +0100, Martin Wilson wrote:
> I am involved in the support of a digital audio editing and production system 
> involving Windows 2k clients storing the audio assets on an AIX backend. The 
> problem we have been experiencing is that we open audio files in the editor 
> we are getting a minimum of 2 sec delay between pressing play and the start 
> of the audio playback. Initial tests show that this delay can be up to 6 
> seconds and that typically there are four to five conversations between the 
> samba server and the win2k DC before the file is begun to be streamed to the 
> client. We haven't established yet whether these conversations are all 
> relevant to the client or whether they are authentication requests for other 
> clients. We are using pass-through authentication. We have noticed that 
> despite our samba config being the default regarding autodisconnect  (ie 
> never disconnect) we see the client application disconnect after a few 
> seconds of inactivity within the audio editor. This leads us to suspect that 
> the application is causing the disconnect by design.  Also, browsing the 
> samba share in windows explorer causes a permanent samba connection, and in 
> this situation we do not have the delay in streaming of the audio.

I need to see a network capture trace of this please.

Thanks,

Jeremy.
--
To unsubscribe from this list go to the following URL and read the
instructions:  https://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issues (compared win2k)

[Isaac Ojeda Llebry]

El Miércoles, 24 de Noviembre de 2004 11:30, escribió:
> Are you sure it is samba who is causing delays and not the file system?

No, you are right. I will try to change the file system from EXT3 to REISER.

> What file system are you using? Is the second access to a file as slow as 
the first?

No, the second access seems faster.

> > socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192 IPTOS_LOWDELAY
> The default values for RCVBUF and SNDBUF are larger, why do you reduce it?

I tried playing with different values from 1024 to 65535 and it makes no 
sensse. The default option of RCVBUF and SNDBUF is 8192


Thanks
--
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issues (compared win2k)

[Gerald Drouillard]

Isaac Ojeda Llebry wrote:
We're running samba in our organization to serve files in a LAN to windows 
machines (almost XP), and we're having some performance issues with small 
files. With big files (ie, ISO images), it works pretty well. But, when small 
files are involved in the transference, problems arise. In the same 
environment, a Windows 2000 machine (the 'old' server) is able to send data 
to clients nearly at double speed.

We've tried to change some parameters to make the performance better but... no 
way.

So, we're looking for any idea or a point to start to search for some 
additional info.

We are using in a Samba-3.0.7-1.3E, Linux AS 3 Update 4 box, with the Red 
Hat's kernel 2.4.21-15.0.3.ELsmp.

Configuration is as follow:
# Global parameters
[global]
workgroup = OURWORKGROUP
server string = Software Server
interfaces = eth1
auth methods = guest, sam, winbind
map to guest = Bad User
null passwords = Yes
guest account = ouruser
passwd program = /usr/bin/passwd %u
passwd chat = *New*UNIX*password* %n\n *ReType*new*UNIX*password* %n\n 
*passwd:*all*authentication*tokens*updated*successfully*
username map = /etc/samba/smbusers
log file = /var/log/samba/smbd.log
max log size = 10240
max xmit = 65535
dns proxy = No
wins server = 192.168.10.10
ldap ssl = no
create mask = 0775
hosts allow = 127., 192.168.
cups options = raw
socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192 
IPTOS_LOWDELAY


Try:
socket options = TCP_NODELAY SO_SNDBUF=65536 SO_RCVBUF=65536 
IPTOS_LOWDELAY
Have a look at:
http://www.drouillard.ca/Tips&Tricks/Samba/Oplocks.htm
--
Regards
--
Gerald Drouillard
Technology Architect
Drouillard & Associates, Inc.
http://www.Drouillard.ca
--
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issue

[eric roseme]

Hi Xiaoqin,
First, if TCP_NODELAY is not being set, that could be your performance 
problem right there.  I have no idea what the problem is with setting 
your socket options.  I guess that you compile your own Samba version, 
so maybe it's time to start investigating your build.

My version of HP CIFS Server on 3.0.5 does not exhibit any of the 
symptoms as seen in your logs.  You can pull down the latest build of 
CIFS 3.0.5 (for testing only) from:

http://www.software.hp.com/portal/swdepot/displayProductInfo.do?productNumber=CIFSTP3
On 2.2 CIFS and Opensource can co-exist on the same system (only one can 
run), but I have not tested this on 3.0 yet.  So you could test with 
CIFS for the socket options to see if you have a build problem.

Second, if you are running opensource, then you are probably calling 
pread/pwrite.  If you are doing that, then you need phlk_28512.  That 
can slow down reads/writes too.

Eric Roseme
Hewlett-Packard
[EMAIL PROTECTED] wrote:
Hi,
In the last a couple of weeks, Eric helped me fixed a couple of my new samba 3.0.5 running on HP-UX 11i hang issues. Right now, people still experience slowness when they run some applications on the samba shares OR recursive list directories on the samba shares. 

There was not a lot of errors in the individual log files. However,there are some 
errors in log.smbd and log.0.0.0.0 file.
1) what is log.0.0.0.0 file? Is it a problem that it exists?
2) In log.smbd file, I saw the following type of errors:
[2004/09/08 09:23:51, 0] lib/util_sock.c:get_peer_addr(978)
  getpeername failed. Error was Invalid argument
[2004/09/08 09:27:52, 0] smbd/server.c:open_sockets_smbd(382)
  open_sockets_smbd: accept: No buffer space available
[2004/09/08 09:30:15, 0] smbd/server.c:open_sockets_smbd(382)
  open_sockets_smbd: accept: No buffer space available
[2004/09/08 09:31:17, 0] lib/util_sock.c:set_socket_options(185)
  Failed to set socket option SO_KEEPALIVE (Error Invalid argument)
[2004/09/08 09:31:17, 0] lib/util_sock.c:set_socket_options(185)
  Failed to set socket option TCP_NODELAY (Error Invalid argument)
[2004/09/08 09:31:17, 0] lib/util_sock.c:get_peer_addr(978)
  getpeername failed. Error was Invalid argument
[2004/09/08 09:33:08, 0] smbd/server.c:open_sockets_smbd(382)
  open_sockets_smbd: accept: No buffer space available
3) In log.0.0.0.0 file, I saw the following type of errors:
[2004/09/08 15:54:22, 0] lib/util_sock.c:get_peer_addr(978)
  getpeername failed. Error was Invalid argument
[2004/09/08 15:54:22, 0] lib/access.c:check_access(326)
[2004/09/08 15:54:22, 0] lib/util_sock.c:get_peer_addr(978)
  getpeername failed. Error was Invalid argument
  Denied connection from  (0.0.0.0)
[2004/09/08 15:54:22, 1] smbd/process.c:process_smb(883)
[2004/09/08 15:54:22, 0] lib/util_sock.c:get_peer_addr(978)
  getpeername failed. Error was Invalid argument
  Connection denied from 0.0.0.0
[2004/09/08 15:54:22, 0] lib/util_sock.c:write_socket_data(413)
  write_socket_data: write failure. Error = Broken pipe
[2004/09/08 15:54:22, 0] lib/util_sock.c:write_socket(437)
  write_socket: Error writing 5 bytes to socket 23: ERRNO = Broken pipe
[2004/09/08 15:54:22, 0] lib/util_sock.c:send_smb(629)
  Error writing 5 bytes to client. -1. (Broken pipe)
Are these real problems and how to get rid of them?
BTW, we have the following configuration in smb.conf file:
  socket options = TCP_NODELAY
Thank you very much for your help!
Xiaoqin Qiu
Agilent Technologies, Inc.
--
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance/stability issue...

[Jeremy Allison]

On Thu, Jul 15, 2004 at 09:01:05PM -0400, Marcello Melfi wrote:
> Hi Jeremy,
> 
> I have added the parameters
> 
>   case sensitive = yes
>   preserve case = no
>   short preserve case = no
>   default case = lower
> 
> to the smb.conf file and everything seems ok now, as far as performance
> goes.
> 
> However, the created files on the share are still in uppercase... I though
> that these settings were to stop the lookup and, as a consequence of the
> default case setting, to have all newly created files in lower case. Any
> idea?

I can't reproduce this with the latest Samba svn code so I'm assuming it's
fixed.

Jeremy
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba performance/stability issue...

[Marcello Melfi]

Hi Jeremy,

Thanks again!

Don't worry, if the performance issue comes back (or was never really
solved), I know someone who I can email to... ;o)

Note: If you need more info on my settings, in case this is a bug, don't
hesitate to ask for it. I'll be glad to help. Remember, I am using Samba
3.0.2a on a Solaris 8 system.

Bye,
 
Marcello

-Original Message-
From: Jeremy Allison [mailto:[EMAIL PROTECTED] 
Sent: July 15, 2004 21:06
To: Marcello Melfi
Cc: 'Jeremy Allison'; [EMAIL PROTECTED]
Subject: Re: [Samba] Samba performance/stability issue...

On Thu, Jul 15, 2004 at 09:01:05PM -0400, Marcello Melfi wrote:
> Hi Jeremy,
> 
> I have added the parameters
> 
>   case sensitive = yes
>   preserve case = no
>   short preserve case = no
>   default case = lower
> 
> to the smb.conf file and everything seems ok now, as far as 
> performance goes.
> 
> However, the created files on the share are still in uppercase... I 
> though that these settings were to stop the lookup and, as a 
> consequence of the default case setting, to have all newly created 
> files in lower case. Any idea?

Ok, could be a bug I'll take a look at it.

Thanks & I'm glad your performance problem seems to be solved !

Jeremy.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance/stability issue...

[Jeremy Allison]

On Thu, Jul 15, 2004 at 09:01:05PM -0400, Marcello Melfi wrote:
> Hi Jeremy,
> 
> I have added the parameters
> 
>   case sensitive = yes
>   preserve case = no
>   short preserve case = no
>   default case = lower
> 
> to the smb.conf file and everything seems ok now, as far as performance
> goes.
> 
> However, the created files on the share are still in uppercase... I though
> that these settings were to stop the lookup and, as a consequence of the
> default case setting, to have all newly created files in lower case. Any
> idea?

Ok, could be a bug I'll take a look at it.

Thanks & I'm glad your performance problem seems to be solved !

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba performance/stability issue...

[Marcello Melfi]

Hi Jeremy,

I have added the parameters

case sensitive = yes
preserve case = no
short preserve case = no
default case = lower

to the smb.conf file and everything seems ok now, as far as performance
goes.

However, the created files on the share are still in uppercase... I though
that these settings were to stop the lookup and, as a consequence of the
default case setting, to have all newly created files in lower case. Any
idea?

Bye,

Marcello

-Original Message-
From: Jeremy Allison [mailto:[EMAIL PROTECTED] 
Sent: July 14, 2004 1:55
To: Marcello Melfi
Cc: 'Jeremy Allison'; [EMAIL PROTECTED]
Subject: Re: [Samba] Samba performance/stability issue...

On Tue, Jul 13, 2004 at 11:06:28PM -0400, Marcello Melfi wrote:
> 
> Regardless of the above, why is it that sometimes I get very good 
> results and many other times bad ones? I would expect that this 
> case-insensitive thing be consistent and therefore always generates 
> bad results. Do you have an explanation for this behavior!

No, sorry, I missed that. If it's a case insensitive problem yes it should
be constant if the files are always being created as new.

> In any case, is there a way to stop this lookup? For example, always 
> use uppercase or something like that...

Yes, set "case sensitive = yes, preserve case = no, short preserve case =
no, default case = lower"
in the share you need this behaviour.

Jeremy.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance/stability issue...

[Jeremy Allison]

On Tue, Jul 13, 2004 at 11:06:28PM -0400, Marcello Melfi wrote:
> 
> Regardless of the above, why is it that sometimes I get very good results
> and many other times bad ones? I would expect that this case-insensitive
> thing be consistent and therefore always generates bad results. Do you have
> an explanation for this behavior!

No, sorry, I missed that. If it's a case insensitive problem yes
it should be constant if the files are always being created as new.

> In any case, is there a way to stop this lookup? For example, always use
> uppercase or something like that...

Yes, set "case sensitive = yes, preserve case = no, short preserve case = no,
default case = lower"
in the share you need this behaviour.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba performance/stability issue...

[Marcello Melfi]

Hi Jeremy,

The C++ application was for testing purposes only. However, the real
application we are using is a commercial package for which I do not have any
control. In real life, that application will run on a few windows
workstations and should export about 3000 to 4000 files per workstation. We
were planning to have one directory per worstation within the samba share.
Can't do it otherwise...

Regardless of the above, why is it that sometimes I get very good results
and many other times bad ones? I would expect that this case-insensitive
thing be consistent and therefore always generates bad results. Do you have
an explanation for this behavior!

In any case, is there a way to stop this lookup? For example, always use
uppercase or something like that...

Note: I really appreciate you for taking the time to respond!

Bye,
 
Marcello
 

-Original Message-
From: Jeremy Allison [mailto:[EMAIL PROTECTED]
Sent: July 13, 2004 21:50
To: Marcello Melfi
Cc: [EMAIL PROTECTED]
Subject: Re: [Samba] Samba performance/stability issue...

On Tue, Jul 13, 2004 at 09:43:02PM -0400, Marcello Melfi wrote:
>  
> I am performing some benchmarks that will reflect the way I am going 
> to use Samba. Basically, I am copying/creating, via a simple C++ 
> program running on the client box, the same 50 K-Bytes file about 
> 10,000 times on the Samba share. Of course, the file is renamed with a 
> sequence number so that at the end there are 10,000 newly created 
> files on the share. As you might have guessed by now, I am not using 
> Samba to simply replace a file server for windows users. I am using 
> Samba so that a windows application (running in the background) can 
> export files to another unix applications. NFS could have been an 
> alternative, but Samba will integrate this export mechanism in a more
transparent fashion.

Are you creating all these files in the same directory ? If so, that's your
answer for why things are slow. Samba has to provide a case-insensitive
lookup for a case-sensitive filesystem. Every time you try and create a file
that doesn't exist Samba has to do a directory scan to see if the file
exists in a different case. This is slow. Fix your app to create into
different directories and you'll find it gets much faster.

Jeremy.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance/stability issue...

[Jeremy Allison]

On Tue, Jul 13, 2004 at 09:43:02PM -0400, Marcello Melfi wrote:
>  
> I am performing some benchmarks that will reflect the way I am going to use
> Samba. Basically, I am copying/creating, via a simple C++ program running on
> the client box, the same 50 K-Bytes file about 10,000 times on the Samba
> share. Of course, the file is renamed with a sequence number so that at the
> end there are 10,000 newly created files on the share. As you might have
> guessed by now, I am not using Samba to simply replace a file server for
> windows users. I am using Samba so that a windows application (running in
> the background) can export files to another unix applications. NFS could
> have been an alternative, but Samba will integrate this export mechanism in
> a more transparent fashion.

Are you creating all these files in the same directory ? If so, that's
your answer for why things are slow. Samba has to provide a case-insensitive
lookup for a case-sensitive filesystem. Every time you try and create a file
that doesn't exist Samba has to do a directory scan to see if the file exists
in a different case. This is slow. Fix your app to create into different
directories and you'll find it gets much faster.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba Performance

[Simon . Harris]

-BEGIN PGP SIGNED MESSAGE-


The contents of this email are intended exclusively for the
addressee. If you are not the addressee you must not read
use or disclose the email contents ; you should notify us
immediately [ by clicking "Reply" ] and delete this email.

Nationwide monitors e-mails to ensure its systems operate
effectively and to minimise the risk of viruses.  Whilst it has
taken reasonable steps to scan this email, it does not
accept liability for any virus that may be contained in it.


Samba 2.2.8a (not upgrade) running on Solaris 9. The system is relatively
new, nothing has changed apart from more data being thrown at it!

When we have certain processes running on the clinet, the a smbd process
gets high CPU and all shares go slow! Files that took seconds now take
minutes. When the process stops the CPU goes down and all returns to normal.
Not sure why such a process is creating this as it's not transferring data
or anything...

A farley simple smb.conf:
# Global parameters
[global]
netbios name = UXHSMSVC01
server string =
encrypt passwords = Yes
username map = /etc/sfw/smbusers
log level = 1
log file = /var/samba/%m.log
show add printer wizard = No
socket address = 10.192.34.61

[mfs_cell_1]
comment = x - Cell 1
path = /mfs/Scan_Cell_1/export
guest account =
valid users = mfssvc
read only = No

[mfs_cell_2]
comment = x - Cell 2
path = /mfs/Scan_Cell_2/export
guest account =
valid users = mfssvc
read only = No

[mfs_cell_3]
comment = x - Cell 3
path = /mfs/Scan_Cell_3/export
guest account =
valid users = mfssvc
read only = No

[nlife]
comment = xxx Life
path = /nlife/export
guest account =
valid users = nlifesvc,svcnlifehsm
read only = No

Turned-up debug to 3 and the log of this client is filled with the following
messages for many thousands of files.

I assume much of this is just samba routines being called/answered but not
sure what is normal...

[2004/01/08 13:59:32, 3] smbd/process.c:(846)
  Transaction 651797 of length 130
[2004/01/08 13:59:32, 3] smbd/process.c:(685)
  switch message SMBntcreateX (pid 25371)
[2004/01/08 13:59:32, 3] smbd/sec_ctx.c:(329)
  setting sec ctx (113, 1) - sec_ctx_stack_ndx = 0
[2004/01/08 13:59:32, 3] smbd/sec_ctx.c:(334)
  1 user groups:
  1
[2004/01/08 13:59:32, 3] smbd/vfs.c:(574)
  vfs_ChDir to /mfs/Scan_Cell_1/export
[2004/01/08 13:59:32, 3] lib/util.c:(387)
  unix_clean_name [/DATA/MORTGAGEA/DRIVE13/V4/indexes/indexxx]
[2004/01/08 13:59:32, 3] smbd/dosmode.c:(111)
  unix_mode(Data/MortgageA/Drive13/V4/INDEXES/INDEXXX) returning 0744
[2004/01/08 13:59:32, 3] lib/util.c:(387)
  unix_clean_name [Data/MortgageA/Drive13/V4/INDEXES/INDEXXX]
[2004/01/08 13:59:32, 2] smbd/open.c:(246)
  mfssvc opened file Data/MortgageA/Drive13/V4/INDEXES/INDEXXX read=Yes
write=N
o (numopen=1)
[2004/01/08 13:59:32, 3] smbd/process.c:(846)
  Transaction 651798 of length 110
[2004/01/08 13:59:32, 3] smbd/process.c:(685)
  switch message SMBtrans2 (pid 25371)
[2004/01/08 13:59:32, 3] smbd/sec_ctx.c:(329)
  setting sec ctx (113, 1) - sec_ctx_stack_ndx = 0
[2004/01/08 13:59:32, 3] smbd/sec_ctx.c:(334)
  1 user groups:
  1
[2004/01/08 13:59:32, 3] smbd/vfs.c:(574)
  vfs_ChDir to /mfs/Scan_Cell_2/export
[2004/01/08 13:59:32, 3] smbd/trans2.c:(1698)
  call_trans2qfilepathinfo: TRANSACT2_QPATHINFO: level = 1004
[2004/01/08 13:59:32, 3] lib/util.c:(387)
  unix_clean_name [/Data/MortgageA/Drive14/V3/DATA]
[2004/01/08 13:59:32, 3] lib/util.c:(387)
  unix_clean_name [Data/MortgageA/Drive14/V3/DATA]
[2004/01/08 13:59:32, 3] smbd/trans2.c:(1730)
  call_trans2qfilepathinfo Data/MortgageA/Drive14/V3/DATA level=1004 call=5
tot
al_data=0
[2004/01/08 13:59:32, 3] smbd/process.c:(846)
  Transaction 651799 of length 113
[2004/01/08 13:59:32, 3] smbd/process.c:(685)
  switch message SMBtrans2 (pid 25371)
[2004/01/08 13:59:32, 3] smbd/sec_ctx.c:(329)
  setting sec ctx (113, 1) - sec_ctx_stack_ndx = 0
[2004/01/08 13:59:32, 3] smbd/sec_ctx.c:(334)
  1 user groups:
  1
[2004/01/08 13:59:32, 3] smbd/vfs.c:(574)
  vfs_ChDir to /mfs/Scan_Cell_1/export
[2004/01/08 13:59:32, 3] smbd/trans2.c:(1698)
  call_trans2qfilepathinfo: TRANSACT2_QPATHINFO: level = 1004
[2004/01/08 13:59:32, 3] lib/util.c:(387)
  unix_clean_name [/DATA/MORTGAGEA/DRIVE13/V4/indexes]
[2004/01/08 13:59:32, 3] lib/util.c:(387)
  unix_clean_name [Data/MortgageA/Drive13/V4/INDEXES]
[2004/01/08 13:59:32, 3] smbd/trans2.c:(1730)
  call_trans2qfilepathinfo Data/MortgageA/Drive13/V4/INDEXES level=1004
call=5
total_data=0
[2004/01/08 13:59:32, 3] smbd/process.c:(846)
  Transaction 651800 of length 125
[2004/01/08 13:5

Re: [Samba] Samba performance on the mainframe

[Jim McDonough]

Samba wrote:

> We too are using Samba as a File Server on our mainframe.  We have
> experienced slowness, but I don't know that it's attributed to Samba
> (atleast in our case).  One thing I do know is I took the same data,
> using the same version of Samba (2.2.7) and installed it on an Intel
> PIII.  The PIII performs better in terms of response time and file
> copy speed.  I hesitate to mention it since I dont' have any hard
> proof or numbers.

> I have yet to narrow down where the problem is, if you have any
> suggestions or want to compare configurations let me know.

How much memory did you give the guest?  That's often the problem.
Linux does so much file caching, so when you're running Linux under VM,
you get multiple levels of OS thinking they're doing the right thing for
performance

Reports of the optimal guest size vary, but I've never heard of anyone
with more than 128MB having good results.  Some folks have found best
results at 44MB.  That's where I'd start.

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba performance on the mainframe

[Samba]

We too 
are using Samba as a File Server on our mainframe.  We have experienced 
slowness, but I don't know that it's attributed to Samba (atleast in our 
case).  One thing I do know is I took the same data, using the same version 
of Samba (2.2.7) and installed it on an Intel PIII.  The PIII performs 
better in terms of response time and file copy speed.  I hesitate to 
mention it since I dont' have any hard proof or numbers.
 
I have 
yet to narrow down where the problem is, if you have any suggestions or want to 
compare configurations let me know.
 
Josh

  -Original Message-From: Sean Angley 
  [mailto:[EMAIL PROTECTED]]Sent: Wednesday, December 11, 2002 2:24 
  PMTo: [EMAIL PROTECTED]Subject: [Samba] Samba 
  performance on the mainframeHey Team.. We notice that 
  repeated file writes to the Samba server slows down under moderate loading. 
  Has anyone else had this fall off in performance and if so, what did you do to 
  correct it?regards.. Sean Angley, 
  P.Eng. [EMAIL PROTECTED]   Host Server 
  Support   ISM Canada    IBM Global 
  Services   One Research Drive   Regina, 
  CANADA  S4S 7H1PHONE:      (306) 790-5199 
  FAX:            (306) 
790-5070

Re: [Samba] Samba Performance question

[Dragan Krnic]

You made my day, Guenther. I believed what they posted on bestbits
that they are still looking for someone to patch Reiser. Have you
actually tried it and it worked?
--

On Mon, 9 Dec 2002 14:56:31   
 Guenther Deschner wrote:
>hi,
>> Such a shame ReiserFS has no ACLs. 
>> By far the best for such situations.
>
>just for the record: 
>
>all the recent SuSE-versions ship with acls on xfs,ext2/ext3,jfs and
>guess what: reiserfs. on SuSE Linux 8.1 you have to enable it though and
>recompile your kernel.


_
Get 25MB, POP3, Spam Filtering with LYCOS MAIL PLUS for $19.95/year.
http://login.mail.lycos.com/brandPage.shtml?pageId=plus&ref=lmtplus
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba Performance question

[Guenther Deschner]

hi,

On Fri, Dec 06, 2002 at 02:20:49PM +0100, Dragan Krnic wrote:
> > ext2/ext3 manipulate the directory entries using lists 
> > so if you have a great many files in one directory you
> > will see performance issues as you describe. The answer 
> > to this is to change filesystem - no mean feat with your
> > data sizes.  Filesystems like XFS and ReiserFS use binary
> > trees to manipulate the directory entries and it is a far
> > faster way of doing things with crowded directories so 
> > you should see an improvemnet.
> 
> Migrating to another file system should not be a big headache, if
> you have the means (free disk space, tapes). Such a shame ReiserFS 
> has no ACLs. By far the best for such situations.

just for the record: 

all the recent SuSE-versions ship with acls on xfs,ext2/ext3,jfs and
guess what: reiserfs. on SuSE Linux 8.1 you have to enable it though and
recompile your kernel.

cheers,
guenther
-- 
Guenther Deschner [EMAIL PROTECTED]
SuSE Linux AGGnuPG: 8EE11688
Berliner Str. 27  phone:  +49 (0) 30 / 430944778
D-13507 Berlin   fax:  +49 (0) 30 / 43732804



msg10282/pgp0.pgp
Description: PGP signature

RE: [Samba] Samba Performance question

[paul . r . schenk]

Hello Wolfgang,

First, to answer your question 1), Samba builds the directory list on the
server.

Now, given what you describe in 2), Samba is not the tool for you. It
handles the situation you describe very poorly.

All is not lost though, since you control the filenames, and presumably the
code that looks for them, I'd rip-out all the samba 8.3-case-insensitive
code in trans2.c (get_lanman2_dir_entry, trans2_find_first, ) and
replace it with Unix opendir and fopen calls etc. This will give you
something that speaks SMB on the wire, but does not implement the
filesystem correctly. It will be fast though.

All the best,
Paul



   
   
  "Belgardt, Wolfgang" 
   
  <[EMAIL PROTECTED] To:  <[EMAIL PROTECTED]>  
   
  m>   cc:  <[EMAIL PROTECTED]>  
   
  Sent by:             Subject: RE: [Samba] Samba Performance 
question
  [EMAIL PROTECTED] 
   
  org  
   
   
   
   
   
  06/12/2002 04:51 PM  
   
   
   
   
   



Hello Paul,

Thanks for your explanations of samba doings in this case.

1) where is samba build the in-memory list?  On server  or on the client? I
think on the server, right?
1a) All file  are 8.3 named files;  installed via a NT client
.
2) To your informations: I believe  the customer software is searching the
files with wildcards. What is the customer doing? He read on a NT client
music CDs and build from every Track on this CD a 30 sec MP3 file.  The
software is  automatic create a 8 character long directory and write  for
all tracks,of the CD,  an mp3 file with an 8.3 name. Only the 3 character
exts is different; to recognize the tracks. The write is not a Problem,
because the time is not relevant for the customer.




Kind Regards / Grüsse

Wolfgang

-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Friday, December 06, 2002 19:33
To: Belgardt, Wolfgang
Cc: [EMAIL PROTECTED]
Subject: Re: [Samba] Samba Performance question


We had the same problem here and I traced it to how Samba pretends to be a
Windows server.

Basically Samba does this:

1) build an in-memory list of a directory's contents, with 8.3 mangled
names
2) When asked for a file, look through the list created in step 1) trying
to find a match.  It tries an exact match and then an 8.3 match.

With large numbers of files in directories (I have one with about 65
files), 1) creates a huge list and 2) takes forever and pegs the CPU at
100%.

In cases with large numbers of files, Windows wins hands-down, because the
8.3-stupid-stuff is handled by the filesystem.

I solved this by making a modified version of a few routines.
1) I make the routines that create the directory list abort after 100 files
and pretend there are no more files.
2) I modified the file opening routine (trans2_readdir, I think) to attempt
to open the file using the filesystem first, bypassing all
case-insenitive-8.3-mangling code. If that fails, I let it try the
look-up-in-a-list method (except for a hard-coded directory where I return
file-not-found if the direct attempt failed).
3) I set 'dont descend' on the big directories, to help users who
mistakenly try to browse the directory with explorer, although mod 1) would
mean they'd only see 100 files anyway.

Making these changes allows my HP9000-D380/2 to outperform a Windows NT 4
Pentium 2 when dealing with directories of over 60 files. Stock Samba
compiled from source (or the depot from itrc) served files from this
directory at about the rate of 5 min/file, with the CPU pegged at 100%. NT
can handle this in less than 1 second. Now I have over 400 people opening
files in this directory all day, and the CPU doesn't even work up a sweat.

The mods I mad

RE: [Samba] Samba Performance question

[Belgardt, Wolfgang]

Hello Paul,

Thanks for your explanations of samba doings in this case.

1) where is samba build the in-memory list?  On server  or on the client? I think on 
the server, right?  
1a) All file  are 8.3 named files;  installed via a NT client
.
2) To your informations: I believe  the customer software is searching the files with 
wildcards. What is the customer doing? He read on a NT client music CDs and build from 
every Track on this CD a 30 sec MP3 file.  The software is  automatic create a 8 
character long directory and write  for all tracks,of the CD,  an mp3 file with an 8.3 
name. Only the 3 character exts is different; to recognize the tracks. The write is 
not a Problem, because the time is not relevant for the customer.  
What is doing with this files? When a customer goes to the CD shop an he will be 
hearing a music from a CD, he put the CD  to a barcode reader and then the software is 
searching via the barcode key the mp3 files to play the music. This search is doing, I 
believe, with wildcards. The time to search is 6 sec  in a directory with 45000 files 
and with a filename with 257 exts.
When we search in this directory a file with fewer exts (ie 15) the search need 3 sec. 
 With much fewer files in directory (1) search is done fewer the 1 sec.



Kind Regards / Grüsse 

Wolfgang 

-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Friday, December 06, 2002 19:33
To: Belgardt, Wolfgang
Cc: [EMAIL PROTECTED]
Subject: Re: [Samba] Samba Performance question


We had the same problem here and I traced it to how Samba pretends to be a
Windows server.

Basically Samba does this:

1) build an in-memory list of a directory's contents, with 8.3 mangled
names
2) When asked for a file, look through the list created in step 1) trying
to find a match.  It tries an exact match and then an 8.3 match.

With large numbers of files in directories (I have one with about 65
files), 1) creates a huge list and 2) takes forever and pegs the CPU at
100%.

In cases with large numbers of files, Windows wins hands-down, because the
8.3-stupid-stuff is handled by the filesystem.

I solved this by making a modified version of a few routines.
1) I make the routines that create the directory list abort after 100 files
and pretend there are no more files.
2) I modified the file opening routine (trans2_readdir, I think) to attempt
to open the file using the filesystem first, bypassing all
case-insenitive-8.3-mangling code. If that fails, I let it try the
look-up-in-a-list method (except for a hard-coded directory where I return
file-not-found if the direct attempt failed).
3) I set 'dont descend' on the big directories, to help users who
mistakenly try to browse the directory with explorer, although mod 1) would
mean they'd only see 100 files anyway.

Making these changes allows my HP9000-D380/2 to outperform a Windows NT 4
Pentium 2 when dealing with directories of over 60 files. Stock Samba
compiled from source (or the depot from itrc) served files from this
directory at about the rate of 5 min/file, with the CPU pegged at 100%. NT
can handle this in less than 1 second. Now I have over 400 people opening
files in this directory all day, and the CPU doesn't even work up a sweat.

The mods I made break what I understand SMB to be. The broken-ness would
only affect old clients (Win 3.1) and clients that try to open 'AFILE.DOC'
and expect to get 'afile.doc'. Since I control what the client requests, I
could get around this. YMMV.

Hope this helps. Does anybody know if changes to address this problem are
in Samba 3?

All the best,
Paul



   
  
  "Belgardt, Wolfgang" 
  
  <[EMAIL PROTECTED] To:  <[EMAIL PROTECTED]>  
  
  m>   cc: 
  
  Sent by: Subject: [Samba] Samba Performance 
question   
  [EMAIL PROTECTED] 
  
  org  
  
   
  
   
 

RE: [Samba] Samba Performance question

[Jim Morris]

Wolfgang,

I think the bottleneck is what another poster has just pointed out - the
fact that Samba is trying to support both long filenames and the older
DOS 8.3 style 'mangled' filenames.  The time spent building the list of
mangled filenames for the huge numbers of files you have must be very
time consuming.  Have you looked at CPU load during the operation that
the customer complains is too slow?  If it is high, that lends credence
to this theory.

The fact that you see a 2X increase in speed when going with ASU seems
to also point to something other than the storage itself as the
problem...
-- 
/---
| Jim Morris  |  Email: [EMAIL PROTECTED]
| |AIM: JFM2001
\---


-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba Performance question

[Belgardt, Wolfgang]

Hello Jim and Noel,

 

Thanks for your helps.

You ask me about
the physical disk configuration.

A short  survey:

The physical Configuration
of the storage is a HP/Compag  SAN fabric
with five HSG80 storage array controllers. The disk on this five controllers
are configured to raid 5 sets. ( All raid 5 sets ( about 25) are summarized  by advfs addvol command to one big advfs
Volume. I believe that we have an very good balancing on the fiber optic SAN
and also on the HSG80 storage controllers.  If I run a search from Tru64,  have excellent time of response.

Parallel we have tested
with ASU (Advanced Server Unix 
Software) and the responds was factor 2 higher as  with samba.

To break down the
large directories was my first  meaning also,  we
have done this depended from the software. Now we can’t more decrease the number
of file in the directories.

I have read in a
document, that samba has changed from  shared memory to use mmap. Is there an  bottleneck maybe? I will look to the
kernel parameters from ubc (unified buffer cache) on Tru64.  I believe it is a buffer problem, but
where?

 

 

 

Kind Regards / Grüsse 

 

Wolfgang 

 

-Original
Message-
From: Jim Morris
[mailto:[EMAIL PROTECTED]]
Sent: Friday, December 06, 2002
17:29
To: Noel Kelly
Cc: Belgardt, Wolfgang;
[EMAIL PROTECTED]
Subject: Re: [Samba] Samba
Performance question

 

On Friday,
December 6, 2002, at 04:13 AM, Noel Kelly wrote: 

 

Someone
else might well know better but 

  

I
believe this is a file system issue.  ext2/ext3 manipulate the directory
entries using lists so if you have a great many files in one directory you will
see performance issues as you describe. 

 

The original
poster is running an HP cluster system with Tru64 v5.1! Linux has nothing to do
with his issues. 

The
answer to this is to change filesystem - no mean feat with your data
sizes.  Filesystems like XFS and ReiserFS use binary trees to manipulate
the directory entries and it is a far faster way of doing things with crowded
directories so you should see an improvemnet. 

 

Probably a
good point, but again, he is limited to the filesystems available under Tru64.
I have only used HP-UX up through V10.0, and am not familiar with Tru64, so
cannot comment on that 

  

I
suppose an alternative short term solution is to get the users to break large
directories up into small ones if the data lends itself to it. 

 

Probably the
best solution - but maybe not what his customer will want to hear 

 

Wolfgang: is
this on a raid array, or some type of other storage array? Could that be the
bottleneck? Is the NT system using comparable storage hardware? 

 

-- 

Jim Morris
([EMAIL PROTECTED]) 

Re: [Samba] Samba Performance question

[paul . r . schenk]

We had the same problem here and I traced it to how Samba pretends to be a
Windows server.

Basically Samba does this:

1) build an in-memory list of a directory's contents, with 8.3 mangled
names
2) When asked for a file, look through the list created in step 1) trying
to find a match.  It tries an exact match and then an 8.3 match.

With large numbers of files in directories (I have one with about 65
files), 1) creates a huge list and 2) takes forever and pegs the CPU at
100%.

In cases with large numbers of files, Windows wins hands-down, because the
8.3-stupid-stuff is handled by the filesystem.

I solved this by making a modified version of a few routines.
1) I make the routines that create the directory list abort after 100 files
and pretend there are no more files.
2) I modified the file opening routine (trans2_readdir, I think) to attempt
to open the file using the filesystem first, bypassing all
case-insenitive-8.3-mangling code. If that fails, I let it try the
look-up-in-a-list method (except for a hard-coded directory where I return
file-not-found if the direct attempt failed).
3) I set 'dont descend' on the big directories, to help users who
mistakenly try to browse the directory with explorer, although mod 1) would
mean they'd only see 100 files anyway.

Making these changes allows my HP9000-D380/2 to outperform a Windows NT 4
Pentium 2 when dealing with directories of over 60 files. Stock Samba
compiled from source (or the depot from itrc) served files from this
directory at about the rate of 5 min/file, with the CPU pegged at 100%. NT
can handle this in less than 1 second. Now I have over 400 people opening
files in this directory all day, and the CPU doesn't even work up a sweat.

The mods I made break what I understand SMB to be. The broken-ness would
only affect old clients (Win 3.1) and clients that try to open 'AFILE.DOC'
and expect to get 'afile.doc'. Since I control what the client requests, I
could get around this. YMMV.

Hope this helps. Does anybody know if changes to address this problem are
in Samba 3?

All the best,
Paul



   
   
  "Belgardt, Wolfgang" 
   
  <[EMAIL PROTECTED] To:  <[EMAIL PROTECTED]>  
   
  m>   cc: 
   
  Sent by: Subject: [Samba] Samba Performance 
question
  [EMAIL PROTECTED] 
   
  org  
   
   
   
   
   
  05/12/2002 04:45 PM  
   
   
   
   
   



Dear all,





I have a difficult Problem with samba 2.2.5, I hope everyone can help me.


My customer has samba 2.2.5 running on a HP Alpha Server ES40 Cluster with
Tru64 V5.1. The share on this Server has  3.1 million files in  16000
directories.


Some one this directories have 45000 files on it.


The problem is: if we try a search a file from this  big directory  via an
NT Client the response time is to large for the the customer.


He has run an similar application on a NT File server. NT responded after 1
sec  and samba need 6 sec.


Can someone explain me what I can do to increase the performance, please?






Kind Regards / Mit freundlichen Grüssen


Wolfgang Belgardt
Customer Support Consultant

Hewlett-Packard GmbH
Customer Support
Bonsiepen 5
D-45136 Essen
Phone: ++49 (0) 201 2663 258
Fax: ++49 (0) 201 2663 200
mobil:   +49 (0171 3357 256)
E-mail:  [EMAIL PROTECTED]
http://www.hp.com/de
__
Hewlett-Packard GmbH
Geschäftsführer: Jörg Menno Harms (Vorsitzender), Jürgen Banhardt, Wolfram
Fischer,
 Rainer Kaczmarczyk, Bärbel Schmidt, Fritz Schuller, Regine Stachelhaus
Vorsitzender des Aufsichtsrats: Heribert Schmitz
Sitz der Gesellschaft: Böblingen, Amtsgericht Böblingen HRB 4081









This message is for the de

Re: [Samba] Samba Performance question

[Jim Morris]

On Friday, December 6, 2002, at 04:13  AM, Noel Kelly wrote:

Someone else might well know better but
 
I believe this is a file system issue.  ext2/ext3 manipulate the directory entries using lists so if you have a great many files in one directory you will see performance issues as you describe.

The original poster is running an HP cluster system with Tru64 v5.1!  Linux has nothing to do with his issues.


The answer to this is to change filesystem - no mean feat with your data sizes.  Filesystems like XFS and ReiserFS use binary trees to manipulate the directory entries and it is a far faster way of doing things with crowded directories so you should see an improvemnet.

Probably a good point, but again, he is limited to the filesystems available under Tru64. I have only used HP-UX up through V10.0, and am not familiar with Tru64, so cannot comment on that
 
I suppose an alternative short term solution is to get the users to break large directories up into small ones if the data lends itself to it.

Probably the best solution - but maybe not what his customer will want to hear


Wolfgang:  is this on a raid array, or some type of other storage array?  Could that be the bottleneck?  Is the NT system using comparable storage hardware?


--
Jim Morris ([EMAIL PROTECTED])

RE: [Samba] Samba Performance question

[Dragan Krnic]

> ext2/ext3 manipulate the directory entries using lists 
> so if you have a great many files in one directory you
> will see performance issues as you describe. The answer 
> to this is to change filesystem - no mean feat with your
> data sizes.  Filesystems like XFS and ReiserFS use binary
> trees to manipulate the directory entries and it is a far
> faster way of doing things with crowded directories so 
> you should see an improvemnet.

Migrating to another file system should not be a big headache, if
you have the means (free disk space, tapes). Such a shame ReiserFS 
has no ACLs. By far the best for such situations.

> I suppose an alternative short term solution is to get
> the users to break large directories up into small ones
> if the data lends itself to it.

Divide and conquer. But maybe there are simpler ways. My
impression is that all transactions are significantly faster
on a Samba than on the comparable Windoze. I'd like to
reproduce your problem. Could you be more specific? Could
you take a representative directory, copy the structure
somewhere else, i.e. just empty files with identical names,
tar it, gzip it and attach it to your mail? Give precisely
the command you're using and the timings you get, network
conditions, platform, version etc. and I'll see what's wrong.

As I said, what turned me on to Samba was when I saw how fast
it could compute a directory size in a property box. With the
copy of the same directory on both sides Samba was beating
Redmond hands down with a factor of at least 10 on any vaguely
non-trivial hierarchy.


_
Get 25MB, POP3, Spam Filtering with LYCOS MAIL PLUS for $19.95/year.
http://login.mail.lycos.com/brandPage.shtml?pageId=plus&ref=lmtplus
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba Performance question

[Noel Kelly]

Someone else might well know better but
 
I 
believe this is a file system issue.  ext2/ext3 manipulate the directory 
entries using lists so if you have a great many files in one directory you will 
see performance issues as you describe.
 
The 
answer to this is to change filesystem - no mean feat with your data 
sizes.  Filesystems like XFS and ReiserFS use binary trees to manipulate 
the directory entries and it is a far faster way of doing things with crowded 
directories so you should see an improvemnet.
 
I 
suppose an alternative short term solution is to get the users to break large 
directories up into small ones if the data lends itself to 
it.
 
HTH
Noel
 
 

  -Original Message-From: Belgardt, Wolfgang 
  [mailto:[EMAIL PROTECTED]]Sent: 05 December 2002 
  20:45To: [EMAIL PROTECTED]Subject: [Samba] Samba 
  Performance question
  
  Dear 
  all,
   
  I have a 
  difficult Problem with samba 2.2.5, I hope everyone can help 
  me.
  My 
  customer has samba 2.2.5 running on a HP Alpha Server ES40 Cluster with Tru64 
  V5.1. The share on this Server has  3.1 million files in  16000 
  directories.
  Some one 
  this directories have 45000 files on it.
  The 
  problem is: if we try a search a file from this  big directory  via an   NT Client the response time is to large 
  for the the customer.
  He has 
  run an similar application on a NT File server. NT responded after 1 sec  and samba need 6 
  sec.
  Can 
  someone explain me what I can do to increase the performance, 
  please?
   
   
  Kind 
  Regards / Mit 
  freundlichen Grüssen 
   
   
  Wolfgang 
  Belgardt
  Customer 
  Support Consultant
   
  Hewlett-Packard 
  GmbH
  Customer 
  Support
  Bonsiepen 
  5
  D-45136 
  Essen
  Phone: 
  ++49 (0) 201 2663 258
  Fax:     ++49 (0) 201 2663 
  200
  mobil: 
    +49 (0171 3357 
  256)
  E-mail: 
   [EMAIL PROTECTED]
  http://www.hp.com/de
  __
  Hewlett-Packard 
  GmbH
  Geschäftsführer: 
  Jörg Menno Harms (Vorsitzender), Jürgen Banhardt, Wolfram 
  Fischer,
   Rainer Kaczmarczyk, Bärbel Schmidt, 
  Fritz Schuller, Regine Stachelhaus
  Vorsitzender 
  des Aufsichtsrats: Heribert Schmitz 
  Sitz 
  der Gesellschaft: Böblingen, Amtsgericht Böblingen HRB 4081
   

RE: [Samba] Samba Performance

[David Brodbeck]

> -Original Message-
> From: mlh [mailto:mlh@;zip.com.au]

> Also, double check that the nic is in fact running full duplex.
> It's best just to force both sides to full duplex since 
> auto-negotiation is so unreliable.

And the really tricky thing about a duplex mismatch is you'll only see
errors in the statistics on one end.  (The end that's running half duplex
will report lots of late collisions.)  The other will report nothing's
wrong.  If one end of the link is an unmanaged switch, there may be *no* way
to diagnose the problem through statistics, short of using a network
sniffer!  BTDT...we have an unmanaged switch that can be set to half or full
duplex on each port, but that does *not* autonegotiate at all.  It's caused
me some real headaches.  The other nasty thing about this problem is short
packets like pings invariably get through fine.  It's only when you try to
copy a big file that the problem crops up.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba Performance

[mlh]

[EMAIL PROTECTED] wrote:



Any ideas where i can get a "things to check" list for performace 
tuning of Samba

Currently copying 300mb of data since 09:00am this morning and still 
going ...now 12:20!

100mb full duplex nic  on a IBM x232 series dual proc Piii-1.2ghz, 
512mb ram, raid 5 - 18.2gb drives

Have you read the Speed.txt and Speed2.txt files that come with Samba?

Also, double check that the nic is in fact running full duplex.
It's best just to force both sides to full duplex since auto-negotiation
is so unreliable.

Matt




--
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba Performance

[Noel Kelly]

Try an 
ifconfig and see if you are getting errors on the NIC first.
 
Noel

  -Original Message-From: [EMAIL PROTECTED] 
  [mailto:[EMAIL PROTECTED]]Sent: 28 October 2002 
  12:38To: [EMAIL PROTECTED]Subject: [Samba] Samba 
  PerformanceAny ideas 
  where i can get a "things to check" list for performace tuning of Samba 
  Currently copying 300mb of data since 
  09:00am this morning and still going ...now 12:20! 100mb full duplex nic  on a IBM x232 series dual 
  proc Piii-1.2ghz, 512mb ram, raid 5 - 18.2gb drives samba -- mandrake 8.2 with 
  2.2.5pre1  according to smbstatus 
  - locks are exclusive+batch Locked 
  files: Pid    DenyMode 
    Access      R/W        Oplock   
          Name -- 
  2276   DENY_ALL   0x30196   
    WRONLY     EXCLUSIVE+BATCH 
   /var/apps/opera/data/s_aentry.cdx   Mon Oct 28 12:28:27 2000 
  any tips would be appreciated 
  kind regards rod 
  ---Incoming mail is certified Virus Free.Checked by 
  AVG anti-virus system (http://www.grisoft.com).Version: 6.0.404 / Virus 
  Database: 228 - Release Date: 
15/10/2002


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.404 / Virus Database: 228 - Release Date: 15/10/2002
 

Re: [Samba] Samba - Performance Issues

[mike]

Hi
I have noticed that win98 clients have better performance tha win2k with sp3
.I have a similar setu with a 800 GB volumes . I will try the same tests
from a windows 98 and compare .Also I have a 4 nic teamed using Intel ANS
softwate that improved performance .
Thanks

- Original Message -
From: <[EMAIL PROTECTED]>
To: "Vinay Kudithipudi" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Wednesday, October 02, 2002 2:24 PM
Subject: Re: [Samba] Samba - Performance Issues


> On Wed, Oct 02, 2002 at 12:57:17PM -0500, Vinay Kudithipudi wrote:
> > Hello Guys,
> >   I am having some problems with configuring SAMBA with regards to
> > performance. We are running SAMBA 2.2.3a on Dual PIII 1Ghz machines with
> > 512MB of RAM. The server is running on a default server installation of
Red
> > Hat 7.2. We have a 500Gb RAID 5 drive using the Promise SX6000 Raid
> > controller.
> >   Currently we are only getting a throughput of ~5MB/S for writes and
> > ~13MB/S for reads. Comparing that to NFS which yealds ~15MB/S for reads
and
> > ~13MB/S for writes. This clearly rules out the Hardware bottleneck since
XFS
> > is able to perform better on the same hardware. Here is the smb.conf I
am
> > using currently
>
> What clients are you using ? This can make a big difference in how
> you tune things.
>
> Jeremy.
> --
> To unsubscribe from this list go to the following URL and read the
> instructions:  http://lists.samba.org/mailman/listinfo/samba
>

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba - Performance Issues

[jra]

On Wed, Oct 02, 2002 at 05:50:49PM -0500, Vinay Kudithipudi wrote:
> Jay - I tried the test without any options (i.e. all default) and still get
> the same results
> 
> Javid - I am using copy
> 
> Jeremy - All clients are Win2k or WinXP.
> 
> I would very much like to blame the hardware for the problem, but since NFS
> yields better performance I am thinking SAMBA may be the cause here. 

I'm not saying it isn't. Are you using PC-NFS between the same clients
and the Samba server to compare ?

Win2K and WinXP clients should be using the large read/write calls, so
larger TCP buffers should be better for them. Do you have any network
traces to look at ?

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba - Performance Issues

[Vinay Kudithipudi]

Jay - I tried the test without any options (i.e. all default) and still get
the same results

Javid - I am using copy

Jeremy - All clients are Win2k or WinXP.

I would very much like to blame the hardware for the problem, but since NFS
yields better performance I am thinking SAMBA may be the cause here. 

Vinay Kudithipudi
Associate Network Operations Engineer
Spirian Technologies Inc.

-Original Message-
From: Jay Ts [mailto:[EMAIL PROTECTED]] 
Sent: Wednesday, October 02, 2002 1:50 PM
To: Vinay Kudithipudi
Cc: [EMAIL PROTECTED]
Subject: Re: [Samba] Samba - Performance Issues


On Wed, Oct 02, 2002 at 12:57:17PM -0500, Vinay Kudithipudi wrote:
> 
> ===SMB.CONF===
> [global]
>   workgroup = MYGROUP
>   netbios name = {HOSTNAME}
>   wins server = {WINSSERVER}
>   server string = {HOSTNAME}
>   security = SHARE
>   encrypt passwords = Yes
>   log file = /var/log/samba/log.%m
>   max log size = 50
>   socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192

IIRC, someone wrote in recently saying that the sizes of SO_RCVBUF and
SO_SNDBUF can have huge effects on performance, and setting them to 8192
(which used to be a good idea) can reduce performance. I suggest removing
them from the socket options and measuring the performance at the defaults,
then try modifying them and comparing performance.

Also, TCP_NODELAY is the default, right?  So maybe just comment out the
socket options parameter, restart the daemons, and check to see if the
problem goes away.

> I was wondering if there is any documentation for fine tuning SAMBA.
> Any help is appreciated. Thanks.

Nowadays, it's usually best to "leave things alone" (i.e., at the defaults).
It's important to not change things in a way that reduces performance.

Jay Ts
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba - Performance Issues

[Jay Ts]

On Wed, Oct 02, 2002 at 12:57:17PM -0500, Vinay Kudithipudi wrote:
> 
> ===SMB.CONF===
> [global]
>   workgroup = MYGROUP
>   netbios name = {HOSTNAME}
>   wins server = {WINSSERVER}
>   server string = {HOSTNAME}
>   security = SHARE
>   encrypt passwords = Yes
>   log file = /var/log/samba/log.%m
>   max log size = 50
>   socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192

IIRC, someone wrote in recently saying that the sizes of SO_RCVBUF
and SO_SNDBUF can have huge effects on performance, and setting
them to 8192 (which used to be a good idea) can reduce performance.
I suggest removing them from the socket options and measuring the
performance at the defaults, then try modifying them and comparing
performance.

Also, TCP_NODELAY is the default, right?  So maybe just comment
out the socket options parameter, restart the daemons, and check
to see if the problem goes away.

> I was wondering if there is any documentation for fine tuning SAMBA. Any
> help is appreciated. Thanks.

Nowadays, it's usually best to "leave things alone" (i.e., at the defaults).
It's important to not change things in a way that reduces performance.

Jay Ts
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba - Performance Issues

[jra]

On Wed, Oct 02, 2002 at 12:57:17PM -0500, Vinay Kudithipudi wrote:
> Hello Guys,
>   I am having some problems with configuring SAMBA with regards to
> performance. We are running SAMBA 2.2.3a on Dual PIII 1Ghz machines with
> 512MB of RAM. The server is running on a default server installation of Red
> Hat 7.2. We have a 500Gb RAID 5 drive using the Promise SX6000 Raid
> controller.
>   Currently we are only getting a throughput of ~5MB/S for writes and
> ~13MB/S for reads. Comparing that to NFS which yealds ~15MB/S for reads and
> ~13MB/S for writes. This clearly rules out the Hardware bottleneck since XFS
> is able to perform better on the same hardware. Here is the smb.conf I am
> using currently

What clients are you using ? This can make a big difference in how
you tune things.

Jeremy.
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba - Performance Issues

[Javid Abdul-AJAVID1]

so are you using xcopy or copy dos command in your script?

-Original Message-
From: Vinay Kudithipudi [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, October 02, 2002 1:21 PM
To: 'Javid Abdul-AJAVID1'; [EMAIL PROTECTED]
Subject: RE: [Samba] Samba - Performance Issues


Javid,

I am running a script which copies a 1Gb files to and from the shared
driver. Am then dividing the time taken by the size of the file. I know it
is very hacked up :), but it should at least give some approximations.
Thanks. 

Vinay Kudithipudi
Associate Network Operations Engineer
Spirian Technologies Inc.

-Original Message-
From: Javid Abdul-AJAVID1 [mailto:[EMAIL PROTECTED]] 
Sent: Wednesday, October 02, 2002 1:13 PM
To: 'Vinay Kudithipudi'; [EMAIL PROTECTED]
Subject: RE: [Samba] Samba - Performance Issues


how are u measuring read and write speeds?

-Original Message-
From: Vinay Kudithipudi [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, October 02, 2002 12:57 PM
To: [EMAIL PROTECTED]
Subject: [Samba] Samba - Performance Issues


Hello Guys,
  I am having some problems with configuring SAMBA with regards to
performance. We are running SAMBA 2.2.3a on Dual PIII 1Ghz machines with
512MB of RAM. The server is running on a default server installation of Red
Hat 7.2. We have a 500Gb RAID 5 drive using the Promise SX6000 Raid
controller.
  Currently we are only getting a throughput of ~5MB/S for writes and
~13MB/S for reads. Comparing that to NFS which yealds ~15MB/S for reads and
~13MB/S for writes. This clearly rules out the Hardware bottleneck since XFS
is able to perform better on the same hardware. Here is the smb.conf I am
using currently

===SMB.CONF===
[global]
workgroup = MYGROUP
netbios name = {HOSTNAME}
wins server = {WINSSERVER}
server string = {HOSTNAME}
security = SHARE
encrypt passwords = Yes
log file = /var/log/samba/log.%m
max log size = 50
socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192
dns proxy = No
oplocks = No
level2 oplocks = No

[homes]
comment = Home Directories
read only = No
browseable = No

[Data]
comment = Data Backup Directory
path = /home/gm/data
guest account = 
valid users = spirian
read only = No
==

I was wondering if there is any documentation for fine tuning SAMBA. Any
help is appreciated. Thanks.

Vinay Kudithipudi
Associate Network Operations Engineer
Spirian Technologies Inc. 
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba - Performance Issues

[Vinay Kudithipudi]

Javid,

I am running a script which copies a 1Gb files to and from the shared
driver. Am then dividing the time taken by the size of the file. I know it
is very hacked up :), but it should at least give some approximations.
Thanks. 

Vinay Kudithipudi
Associate Network Operations Engineer
Spirian Technologies Inc.

-Original Message-
From: Javid Abdul-AJAVID1 [mailto:[EMAIL PROTECTED]] 
Sent: Wednesday, October 02, 2002 1:13 PM
To: 'Vinay Kudithipudi'; [EMAIL PROTECTED]
Subject: RE: [Samba] Samba - Performance Issues


how are u measuring read and write speeds?

-Original Message-
From: Vinay Kudithipudi [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, October 02, 2002 12:57 PM
To: [EMAIL PROTECTED]
Subject: [Samba] Samba - Performance Issues


Hello Guys,
  I am having some problems with configuring SAMBA with regards to
performance. We are running SAMBA 2.2.3a on Dual PIII 1Ghz machines with
512MB of RAM. The server is running on a default server installation of Red
Hat 7.2. We have a 500Gb RAID 5 drive using the Promise SX6000 Raid
controller.
  Currently we are only getting a throughput of ~5MB/S for writes and
~13MB/S for reads. Comparing that to NFS which yealds ~15MB/S for reads and
~13MB/S for writes. This clearly rules out the Hardware bottleneck since XFS
is able to perform better on the same hardware. Here is the smb.conf I am
using currently

===SMB.CONF===
[global]
workgroup = MYGROUP
netbios name = {HOSTNAME}
wins server = {WINSSERVER}
server string = {HOSTNAME}
security = SHARE
encrypt passwords = Yes
log file = /var/log/samba/log.%m
max log size = 50
socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192
dns proxy = No
oplocks = No
level2 oplocks = No

[homes]
comment = Home Directories
read only = No
browseable = No

[Data]
comment = Data Backup Directory
path = /home/gm/data
guest account = 
valid users = spirian
read only = No
==

I was wondering if there is any documentation for fine tuning SAMBA. Any
help is appreciated. Thanks.

Vinay Kudithipudi
Associate Network Operations Engineer
Spirian Technologies Inc. 
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

RE: [Samba] Samba - Performance Issues

[Javid Abdul-AJAVID1]

how are u measuring read and write speeds?

-Original Message-
From: Vinay Kudithipudi [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, October 02, 2002 12:57 PM
To: [EMAIL PROTECTED]
Subject: [Samba] Samba - Performance Issues


Hello Guys,
  I am having some problems with configuring SAMBA with regards to
performance. We are running SAMBA 2.2.3a on Dual PIII 1Ghz machines with
512MB of RAM. The server is running on a default server installation of Red
Hat 7.2. We have a 500Gb RAID 5 drive using the Promise SX6000 Raid
controller.
  Currently we are only getting a throughput of ~5MB/S for writes and
~13MB/S for reads. Comparing that to NFS which yealds ~15MB/S for reads and
~13MB/S for writes. This clearly rules out the Hardware bottleneck since XFS
is able to perform better on the same hardware. Here is the smb.conf I am
using currently

===SMB.CONF===
[global]
workgroup = MYGROUP
netbios name = {HOSTNAME}
wins server = {WINSSERVER}
server string = {HOSTNAME}
security = SHARE
encrypt passwords = Yes
log file = /var/log/samba/log.%m
max log size = 50
socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192
dns proxy = No
oplocks = No
level2 oplocks = No

[homes]
comment = Home Directories
read only = No
browseable = No

[Data]
comment = Data Backup Directory
path = /home/gm/data
guest account = 
valid users = spirian
read only = No
==

I was wondering if there is any documentation for fine tuning SAMBA. Any
help is appreciated. Thanks.

Vinay Kudithipudi
Associate Network Operations Engineer
Spirian Technologies Inc. 
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba
-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issues

[John Coston]

Hi all
just an update on our situation. We are not having any issues today, 
and currently have about 40 users with about 120 connections listed by 
smbstatus, and the load average is under 1%. We haven't seen any of the 
smbd processes running away with the CPU so far, fingers crossed. The 
primary change we made was implementing roaming profiles for our users, 
and disabling the mandatory profile. My suspicion is that while 
mandatory profiles were in place, we had some locking problems which 
resulted in the smbd processes utilizing all of the CPU. After shutting 
down completely, and manually killing off the hung processes, things 
are going much better. We are still running without the patch for 
locking issues that Robert mentions below, but that will be rectified 
soon.
We are going to test out the file locking/mandatory profile hypothesis 
with our backup machine in the next few days, and hopefully will be 
able to provide some more detailed information on exactly what is 
happening.
Thanks to all for the guidance, help & suggestions.
John

On Tuesday, September 24, 2002, at 07:06  PM, Robert Stuart wrote:

> Hi,
>
> I'm rather interested in the outcome of this on or off the list; but I
> suspect there will be other people on the list who are interested -
> please keep posting to the list :-)
>
> I think we have very similar HW.  We have a dual CPU (1.4G PIII) 
> LPr2000
> netserver with 10k and 15k drives.  We normally sit around < 5% cpu so
> it seems to be something about your config, probably not your HW.  Our
> box also does internal http, mail, dns, fax, lpd, ldap etc.
>
> Can I suggest you run top and see what seems to be using CPU time?  
> Does
> %CPU in the process list include sys time?  I think it does, if so, it
> will help you get an idea of what is contributing to the 60% sys time
> (even if top doesn't include sys then it is likely to be showing
> culprits anyhow).  Do you run anything else on this machine (eg
> oracle)?  :-)  Perhaps posting  ps axf and a copy of a top page might
> help.
>
> Your LDAP backend... is it getting busy?  Are the relevant things
> indexed?
>
> I'd upgrade your kernel to the last RH7.3 errata (2.4.18-10?).
>
> You will likely get locking issues (discussed on samba-technical show
> stopper) if you are using ldap sam, I patched the samba 2.2.5 rawhide
> rpm.  I'll send that in a separate email to you.
>
> Good luck.
>
> Robert Stuart
> Systems Administrator
> Ph: 61 7 3864 0364
> Fax: 61 7 3221 2553
> --
> To unsubscribe from this list go to the following URL and read the
> instructions:  http://lists.samba.org/mailman/listinfo/samba
>

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issues

[Robert Stuart]

Hi,

I'm rather interested in the outcome of this on or off the list; but I
suspect there will be other people on the list who are interested -
please keep posting to the list :-)

I think we have very similar HW.  We have a dual CPU (1.4G PIII) LPr2000
netserver with 10k and 15k drives.  We normally sit around < 5% cpu so
it seems to be something about your config, probably not your HW.  Our
box also does internal http, mail, dns, fax, lpd, ldap etc.

Can I suggest you run top and see what seems to be using CPU time?  Does
%CPU in the process list include sys time?  I think it does, if so, it
will help you get an idea of what is contributing to the 60% sys time
(even if top doesn't include sys then it is likely to be showing
culprits anyhow).  Do you run anything else on this machine (eg
oracle)?  :-)  Perhaps posting  ps axf and a copy of a top page might
help.

Your LDAP backend... is it getting busy?  Are the relevant things
indexed?

I'd upgrade your kernel to the last RH7.3 errata (2.4.18-10?).

You will likely get locking issues (discussed on samba-technical show
stopper) if you are using ldap sam, I patched the samba 2.2.5 rawhide
rpm.  I'll send that in a separate email to you.

Good luck.

Martin MOKREJ© wrote:
> 
> On 24 Sep 2002, Bradley W. Langhorst wrote:
> 
> > On Tue, 2002-09-24 at 18:07, Martin MOKREJ© wrote:
> > > On 24 Sep 2002, Bradley W. Langhorst wrote:
> > >
> > > > probably you want to run the iostat 1 during heavier load...
> > > > however the summary result does look funny to me...
> > > >
> > > > On my system we have ~ 1:1 ratio of reads to writes
> > > > you have a ~ 1:200 ratio of reads to writes.
> > > > Does that make sense in your environment?
> > >
> > > Didn't you have a look into the first lines where's the summary output
> > > from iostat?
> > >
> > sure i did... maybe i'm misreading it
> > avg-cpu:  %user   %nice%sys   %idle
> >   16.790.00   26.39   56.82
> >
> > Device:tps   Blk_read/s   Blk_wrtn/s   Blk_read   Blk_wrtn
> > dev8-0   11.80 2.24   184.95 541354   44609586
> > dev8-1   11.80 1.67   184.95 403090   44609586
> >
> > Unless I'm mistaken the summary says that (on average) he only has 2
> > reads/s vs 185 writes/s
> 
> You are right, sure. But my point was why is the *current* load that high.
> That's why I had a look on statistics per every second.
> 
> >
> > > All those other remaining output lines show zero disk
> > > activity ...
> > That's why I suggested he run iostat when the system is under more load.
> 
> In principal I agree anyway. ;)
> 
> >
> >
> > > >
> > > > It doesn't look like your system has a memory problem so i'd not worry
> > > > about vmstat.
> > >
> > > I'd would worry. Actually, what says "dmesg" command?
> > The reason I think memory is no problem here is:
> > Mem:  3229040K av, 3166372K used,   62668K free,   0K shrd,  148480K
> > buff
> > Swap:  513976K av,   0K used,  513976K free 2758060K
> > cached
> >
> > so the swap file has not been touched and there is 2.7gig of disk being
> > cached in RAM
> 
> Yes, but the "sy" column in the output of vmstat was quite high. That
> worries me. ;)
> 
> I think we should not Cc: the samba email list to these emails anymore. ;)
> --
> Martin Mokrejs <[EMAIL PROTECTED]>, <[EMAIL PROTECTED]>
> PGP5.0i key is at http://www.natur.cuni.cz/~mmokrejs
> MIPS / Institute for Bioinformatics 
> GSF - National Research Center for Environment and Health
> Ingolstaedter Landstrasse 1, D-85764 Neuherberg, Germany
> tel.: +49-89-3187 3683 , fax: +49-89-3187 3585
> 
> --
> To unsubscribe from this list go to the following URL and read the
> instructions:  http://lists.samba.org/mailman/listinfo/samba

-- 
Robert Stuart
Systems Administrator
Ph: 61 7 3864 0364
Fax: 61 7 3221 2553
--
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issues

[Martin MOKREJŠ]

On 24 Sep 2002, Bradley W. Langhorst wrote:

> On Tue, 2002-09-24 at 18:07, Martin MOKREJŠ wrote:
> > On 24 Sep 2002, Bradley W. Langhorst wrote:
> >
> > > probably you want to run the iostat 1 during heavier load...
> > > however the summary result does look funny to me...
> > >
> > > On my system we have ~ 1:1 ratio of reads to writes
> > > you have a ~ 1:200 ratio of reads to writes.
> > > Does that make sense in your environment?
> >
> > Didn't you have a look into the first lines where's the summary output
> > from iostat?
> >
> sure i did... maybe i'm misreading it
> avg-cpu:  %user   %nice%sys   %idle
>   16.790.00   26.39   56.82
>
> Device:tps   Blk_read/s   Blk_wrtn/s   Blk_read   Blk_wrtn
> dev8-0   11.80 2.24   184.95 541354   44609586
> dev8-1   11.80 1.67   184.95 403090   44609586
>
> Unless I'm mistaken the summary says that (on average) he only has 2
> reads/s vs 185 writes/s

You are right, sure. But my point was why is the *current* load that high.
That's why I had a look on statistics per every second.

>
> > All those other remaining output lines show zero disk
> > activity ...
> That's why I suggested he run iostat when the system is under more load.

In principal I agree anyway. ;)

>
>
> > >
> > > It doesn't look like your system has a memory problem so i'd not worry
> > > about vmstat.
> >
> > I'd would worry. Actually, what says "dmesg" command?
> The reason I think memory is no problem here is:
> Mem:  3229040K av, 3166372K used,   62668K free,   0K shrd,  148480K
> buff
> Swap:  513976K av,   0K used,  513976K free 2758060K
> cached
>
> so the swap file has not been touched and there is 2.7gig of disk being
> cached in RAM

Yes, but the "sy" column in the output of vmstat was quite high. That
worries me. ;)

I think we should not Cc: the samba email list to these emails anymore. ;)
-- 
Martin Mokrejs <[EMAIL PROTECTED]>, <[EMAIL PROTECTED]>
PGP5.0i key is at http://www.natur.cuni.cz/~mmokrejs
MIPS / Institute for Bioinformatics 
GSF - National Research Center for Environment and Health
Ingolstaedter Landstrasse 1, D-85764 Neuherberg, Germany
tel.: +49-89-3187 3683 , fax: +49-89-3187 3585

-- 
To unsubscribe from this list go to the following URL and read the
instructions:  http://lists.samba.org/mailman/listinfo/samba

Re: [Samba] Samba performance issues

[John Coston]

Here is the dmesg file:

Linux version 2.4.18-3 ([EMAIL PROTECTED]) (gcc version 
2.96 2731 (Red Hat Linux 7.3 2.96-110)) #1 Thu Apr 18 07:37:53 EDT 
2002
BIOS-provided physical RAM map:
  BIOS-e820:  - 0009ec00 (usable)
  BIOS-e820: 0009ec00 - 000a (reserved)
  BIOS-e820: 000e9400 - 0010 (reserved)
  BIOS-e820: 0010 - c7ff (usable)
  BIOS-e820: c7ff - c7fffc00 (ACPI data)
  BIOS-e820: c7fffc00 - c800 (ACPI NVS)
  BIOS-e820: fec0 - fec1 (reserved)
  BIOS-e820: fee0 - fee01000 (reserved)
  BIOS-e820: fff8 - 0001 (reserved)
2303MB HIGHMEM available.
On node 0 totalpages: 819184
zone(0): 4096 pages.
zone(1): 225280 pages.
zone(2): 589808 pages.
Kernel command line: ro root=/dev/md4
Initializing CPU#0
Detected 1000.092 MHz processor.
Console: colour VGA+ 80x25
Calibrating delay loop... 1992.29 BogoMIPS
Memory: 3228532k/3276736k available (1119k kernel code, 47812k 
reserved, 775k data, 280k init, 2359232k highmem)
Dentry cache hash table entries: 262144 (order: 9, 2097152 bytes)
Inode cache hash table entries: 262144 (order: 9, 2097152 bytes)
Mount-cache hash table entries: 65536 (order: 7, 524288 bytes)
Buffer cache hash table entries: 262144 (order: 8, 1048576 bytes)
Page-cache hash table entries: 524288 (order: 9, 2097152 bytes)
CPU: Before vendor init, caps: 0383fbff  , vendor = 0
CPU: L1 I cache: 16K, L1 D cache: 16K
CPU: L2 cache: 256K
CPU: After vendor init, caps: 0383fbff   
Intel machine check architecture supported.
Intel machine check reporting enabled on CPU#0.
CPU: After generic, caps: 0383fbff   
CPU: Common caps: 0383fbff   
CPU: Intel Pentium III (Coppermine) stepping 0a
Enabling fast FPU save and restore... done.
Enabling unmasked SIMD FPU exception support... done.
Checking 'hlt' instruction... OK.
POSIX conformance testing by UNIFIX
mtrr: v1.40 (20010327) Richard Gooch ([EMAIL PROTECTED])
mtrr: detected mtrr type: Intel
PCI: PCI BIOS revision 2.10 entry at 0xfda11, last bus=1
PCI: Using configuration type 1
PCI: Probing PCI hardware
PCI: Discovered primary peer bus 01 [IRQ]
isapnp: Scanning for PnP cards...
isapnp: No Plug & Play device found
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
apm: BIOS not found.
Starting kswapd
allocated 64 pages and 64 bhs reserved for the highmem bounces
VFS: Diskquotas version dquot_6.5.0 initialized
pty: 2048 Unix98 ptys configured
Serial driver version 5.05c (2001-07-08) with MANY_PORTS MULTIPORT 
SHARE_IRQ SERIAL_PCI ISAPNP enabled
ttyS00 at 0x03f8 (irq = 4) is a 16550A
Real Time Clock Driver v1.10e
block: 1024 slots per queue, batch=256
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 33MHz system bus speed for PIO modes; override with 
idebus=xx
SvrWks OSB4: IDE controller on PCI bus 00 dev 79
SvrWks OSB4: chipset revision 0
SvrWks OSB4: not 100% native mode: will probe irqs later
 ide0: BM-DMA at 0x1800-0x1807, BIOS settings: hda:pio, hdb:pio
 ide1: BM-DMA at 0x1808-0x180f, BIOS settings: hdc:pio, hdd:pio
hda: CD-224E, ATAPI CD/DVD-ROM drive
hdc: no response (status = 0xa1), resetting drive
hdc: no response (status = 0xa1)
hdd: no response (status = 0xa1), resetting drive
hdd: no response (status = 0xa1)
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide-floppy driver 0.99.newide
Floppy drive(s): fd0 is 1.44M
FDC 0 is a National Semiconductor PC87306
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
ide-floppy driver 0.99.newide
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 32768 buckets, 256Kbytes
TCP: Hash tables configured (established 262144 bind 65536)
Linux IP multicast router 0.06 plus PIM-SM
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
RAMDISK: Compressed image found at block 0
Freeing initrd memory: 224k freed
VFS: Mounted root (ext2 filesystem).
SCSI subsystem driver Revision: 1.00
kmod: failed to exec /sbin/modprobe -s -k scsi_hostadapter, errno = 2
sym53c8xx: at PCI bus 1, device 5, function 0
sym53c8xx: 53c1010-33 detected with Symbios NVRAM
sym53c8xx: at PCI bus 1, device 5, function 1
sym53c8xx: 53c1010-33 detected with Symbios NVRAM
sym53c1010-33-0: rev 0x1 on pci bus 1 device 5 function 0 irq 5
sym53c1010-33-0: Symbios format NVRAM, ID 7, Fast-80, Parity Checking
sym53c1010-33-0: on-chip RAM at 0xfd004000
sym53c1010-33-0: restart (scsi reset).
sym53c1010-33-0: handling phase mismatch from SCRIPTS.
sym53c1010-33-0: Downloading SCSI SCRIPTS.
sym53c1010-33-1: rev 0x1 on pci bus 1 device 5 function 1 irq 9
sym53c1010-33-1: Symbios format NVRAM, ID 7, Fast-80, Pa