Re: [Samba] Aplication slow after migration
Scott Lovenberg escribió:
Felipe Martinez Hermo wrote:
Scott Lovenberg escribió:
On Feb 6, 2008 4:19 AM, Felipe Martinez Hermo
<[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>> wrote:
Sinisa Bandin escribió:
>
>
> Felipe Martinez Hermo wrote:
>>
>>>> OK, so we're apples to apples, so to speak; the servers are
tuned
>>>> the same. I'll assume your disks are tuned from hdparm and
up to
>>>> snuff, otherwise you wouldn't be tuning sockets ;). Did
your old
>>>> server have samba settings for oplocks set?
>>>>
>>>>
>>>> --
>>>> Peace and Blessings,
>>>> -Scott.
>>>>
>>>> "Of course, that's just my opinion; I could be wrong"
>>>> -Dennis Miller
>>> Erm, sorry, I didn't catch that you had 2 .conf files
there. I'm
>>> back to the drawing board. Sorry about that. Anyone else
have any
>>> ideas?
>> Yes, that's whats shocking me. Apparently we're apples to
apples.
>> Except for the kernel (new&slow 2.6.18-4-686 vs old&fast 2.6.8)
>>
>> I've sniffed both eth0 interfaces and I've got some more
information.
>> When talking to the slow server, the client needs to send 76
"TCP
>> segment of a reassembled PDU" that are not sent when talking
to the
>> old and fast server.
>>
>> How can I workaround this issue? Should I lower server's MTU?
How much?
>>
>> Thank you
> Do you happen to have a Realtek 8169 based gigabit ethernet in
new
> server?
>
> If you do, I had the same problem several times last year, and
solved
> all of them by changing motherboards (all were integrated, and I
like
> them to stay that way because I can achieve full gigabit speed
with
> several concurent clients)
>
> Best regards,
> Sinisa Bandin
>
>
No, machines are out-of-the-box HP DL servers:
Ethernet controller: Broadcom Corporation NetXtreme BCM5705_2
Gigabit
Ethernet (rev 03)
I've made a spreadsheet with summarizing wireshark results and
comparing
results for both servers. You can see it here:
http://spreadsheets.google.com/ccc?key=pnLL2fInqFq2YKuZIphtQdA
It's meaningful that fast server makes 406 Trans2 calls, while slow
server makes 616 calls to perform the same operation. The
difference is
mainly in QUERY_PATH_INFO (200 vs 305) and FIND_FIRST2 (94 vs 199)
calls.
Next try: change ethernet wire? :-?
--
==
Felipe Martínez Hermo
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
==
Servicios Informáticos
UGT Galicia
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
==
--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/listinfo/samba
Hrm, are you using SACKs or DSACKs or tcp_low_delay in
/proc/sys/net/somethingOrOther? They didn't change congestion
control default in your upstream kernel, did they? Should be "reno"
by default. Doing a netstat -a, do you have many packets queued in
either direction? This one is puzzling me.
--
Peace and Blessings,
-Scott.
Apparently everything is configured the same way in /proc/sys/net
(both sack & dsack = 1). Regarding the kernel, Old&fast kernel is
2.6.8 (no /proc/sys/net/ipv4/tcp_congestion_control) while new&slow
is 2.6.18-4-686 and congestion control is bic:
[EMAIL PROTECTED]:~$ cat /proc/sys/net/ipv4/tcp_congestion_control
bic
Should I try other congestion control algorithm?
I've made this rudimentary test, and old server is a little bit
faster, but I don't know if it is meaningful at all.
[EMAIL PROTECTED]:~$ ping -i 0.2 fast_server --- fast_server ping
statistics ---
2156 packets transmitted, 2156 received, 0% packet loss, time 431208ms
rtt min/avg/max/mdev = 0.135/0.171/0.245/0.018 ms
[EMAIL PROTECTED]:~$ ping -i 0.2 slow_server
--- slow_server ping statistics ---
2146 packets transmitted, 2146 received, 0% packet loss, time 429165ms
rtt min/avg/max/mdev = 0.152/0.179/0.333/0.021 ms
Regards,
try:
echo "reno" > /proc/sys/net/ipv4/tcp_congestion_control
That'll make sure the tcp/ip stack isn't messing with the tests by
doing window scaling and such. OK, that's one more variable
isolated... let's see what happens. Sorry that this i
Re: Re: [Samba] Aplication slow after migration
Scott Lovenberg escribió:
On Feb 6, 2008 4:19 AM, Felipe Martinez Hermo <[EMAIL PROTECTED]
<mailto:[EMAIL PROTECTED]>> wrote:
Sinisa Bandin escribió:
>
>
> Felipe Martinez Hermo wrote:
>>
>>>> OK, so we're apples to apples, so to speak; the servers are tuned
>>>> the same. I'll assume your disks are tuned from hdparm and up to
>>>> snuff, otherwise you wouldn't be tuning sockets ;). Did your old
>>>> server have samba settings for oplocks set?
>>>>
>>>>
>>>> --
>>>> Peace and Blessings,
>>>> -Scott.
>>>>
>>>> "Of course, that's just my opinion; I could be wrong"
>>>> -Dennis Miller
>>> Erm, sorry, I didn't catch that you had 2 .conf files there. I'm
>>> back to the drawing board. Sorry about that. Anyone else
have any
>>> ideas?
>> Yes, that's whats shocking me. Apparently we're apples to apples.
>> Except for the kernel (new&slow 2.6.18-4-686 vs old&fast 2.6.8)
>>
>> I've sniffed both eth0 interfaces and I've got some more
information.
>> When talking to the slow server, the client needs to send 76 "TCP
>> segment of a reassembled PDU" that are not sent when talking to the
>> old and fast server.
>>
>> How can I workaround this issue? Should I lower server's MTU?
How much?
>>
>> Thank you
> Do you happen to have a Realtek 8169 based gigabit ethernet in new
> server?
>
> If you do, I had the same problem several times last year, and
solved
> all of them by changing motherboards (all were integrated, and I
like
> them to stay that way because I can achieve full gigabit speed with
> several concurent clients)
>
> Best regards,
> Sinisa Bandin
>
>
No, machines are out-of-the-box HP DL servers:
Ethernet controller: Broadcom Corporation NetXtreme BCM5705_2 Gigabit
Ethernet (rev 03)
I've made a spreadsheet with summarizing wireshark results and
comparing
results for both servers. You can see it here:
http://spreadsheets.google.com/ccc?key=pnLL2fInqFq2YKuZIphtQdA
It's meaningful that fast server makes 406 Trans2 calls, while slow
server makes 616 calls to perform the same operation. The
difference is
mainly in QUERY_PATH_INFO (200 vs 305) and FIND_FIRST2 (94 vs 199)
calls.
Next try: change ethernet wire? :-?
--
==
Felipe Martínez Hermo
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
==
Servicios Informáticos
UGT Galicia
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
==
--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/listinfo/samba
Hrm, are you using SACKs or DSACKs or tcp_low_delay in
/proc/sys/net/somethingOrOther? They didn't change congestion control
default in your upstream kernel, did they? Should be "reno" by
default. Doing a netstat -a, do you have many packets queued in
either direction? This one is puzzling me.
--
Peace and Blessings,
-Scott.
Apparently everything is configured the same way in /proc/sys/net (both
sack & dsack = 1). Regarding the kernel, Old&fast kernel is 2.6.8 (no
/proc/sys/net/ipv4/tcp_congestion_control) while new&slow is
2.6.18-4-686 and congestion control is bic:
[EMAIL PROTECTED]:~$ cat /proc/sys/net/ipv4/tcp_congestion_control
bic
Should I try other congestion control algorithm?
I've made this rudimentary test, and old server is a little bit faster,
but I don't know if it is meaningful at all.
[EMAIL PROTECTED]:~$ ping -i 0.2 fast_server
--- fast_server ping statistics ---
2156 packets transmitted, 2156 received, 0% packet loss, time 431208ms
rtt min/avg/max/mdev = 0.135/0.171/0.245/0.018 ms
[EMAIL PROTECTED]:~$ ping -i 0.2 slow_server
--- slow_server ping statistics ---
2146 packets transmitted, 2146 received, 0% packet loss, time 429165ms
rtt min/avg/max/mdev = 0.152/0.179/0.333/0.021 ms
Regards,
--
==
Felipe Martínez Hermo
[EMAIL PROTECTED]
[EMAIL PROTECTED]
==
Servicios Informáticos
UGT Galicia
[EMAIL PROTECTED]
[EMAIL PROTECTED]
==
--
To unsubscribe from this list go to the following URL and read the
instructions: https://lists.samba.org/mailman/listinfo/samba
Re: [Samba] Aplication slow after migration
Sinisa Bandin escribió: Felipe Martinez Hermo wrote: OK, so we're apples to apples, so to speak; the servers are tuned the same. I'll assume your disks are tuned from hdparm and up to snuff, otherwise you wouldn't be tuning sockets ;). Did your old server have samba settings for oplocks set? -- Peace and Blessings, -Scott. "Of course, that's just my opinion; I could be wrong" -Dennis Miller Erm, sorry, I didn't catch that you had 2 .conf files there. I'm back to the drawing board. Sorry about that. Anyone else have any ideas? Yes, that's whats shocking me. Apparently we're apples to apples. Except for the kernel (new&slow 2.6.18-4-686 vs old&fast 2.6.8) I've sniffed both eth0 interfaces and I've got some more information. When talking to the slow server, the client needs to send 76 "TCP segment of a reassembled PDU" that are not sent when talking to the old and fast server. How can I workaround this issue? Should I lower server's MTU? How much? Thank you Do you happen to have a Realtek 8169 based gigabit ethernet in new server? If you do, I had the same problem several times last year, and solved all of them by changing motherboards (all were integrated, and I like them to stay that way because I can achieve full gigabit speed with several concurent clients) Best regards, Sinisa Bandin No, machines are out-of-the-box HP DL servers: Ethernet controller: Broadcom Corporation NetXtreme BCM5705_2 Gigabit Ethernet (rev 03) I've made a spreadsheet with summarizing wireshark results and comparing results for both servers. You can see it here: http://spreadsheets.google.com/ccc?key=pnLL2fInqFq2YKuZIphtQdA It's meaningful that fast server makes 406 Trans2 calls, while slow server makes 616 calls to perform the same operation. The difference is mainly in QUERY_PATH_INFO (200 vs 305) and FIND_FIRST2 (94 vs 199) calls. Next try: change ethernet wire? :-? -- == Felipe Martínez Hermo [EMAIL PROTECTED] [EMAIL PROTECTED] == Servicios Informáticos UGT Galicia [EMAIL PROTECTED] [EMAIL PROTECTED] == -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
Re: [Samba] Aplication slow after migration
OK, so we're apples to apples, so to speak; the servers are tuned the same. I'll assume your disks are tuned from hdparm and up to snuff, otherwise you wouldn't be tuning sockets ;). Did your old server have samba settings for oplocks set? -- Peace and Blessings, -Scott. "Of course, that's just my opinion; I could be wrong" -Dennis Miller Erm, sorry, I didn't catch that you had 2 .conf files there. I'm back to the drawing board. Sorry about that. Anyone else have any ideas? Yes, that's whats shocking me. Apparently we're apples to apples. Except for the kernel (new&slow 2.6.18-4-686 vs old&fast 2.6.8) I've sniffed both eth0 interfaces and I've got some more information. When talking to the slow server, the client needs to send 76 "TCP segment of a reassembled PDU" that are not sent when talking to the old and fast server. How can I workaround this issue? Should I lower server's MTU? How much? Thank you -- == Felipe Martínez Hermo [EMAIL PROTECTED] [EMAIL PROTECTED] == Servicios Informáticos UGT Galicia [EMAIL PROTECTED] [EMAIL PROTECTED] == -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
Re: [Samba] Aplication slow after migration
I've got vfs_cache_pressure = 100 on both servers and ext3 filesystems on both. These are the files on /proc/sys/vm: slow server: file value -rw-r--r-- 1 root root 0 2008-02-01 13:31 block_dump -rw-r--r-- 1 root root 0 2008-02-01 13:31 dirty_background_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:31 dirty_expire_centisecs -rw-r--r-- 1 root root 0 2008-02-01 13:31 dirty_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:31 dirty_writeback_centisecs -rw-r--r-- 1 root root 0 2008-02-01 13:31 drop_caches -rw-r--r-- 1 root root 0 2008-02-01 13:31 laptop_mode -rw-r--r-- 1 root root 0 2008-02-01 13:31 legacy_va_layout -rw-r--r-- 1 root root 0 2008-02-01 13:31 lowmem_reserve_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:31 max_map_count -rw-r--r-- 1 root root 0 2008-02-01 13:31 min_free_kbytes -r--r--r-- 1 root root 0 2008-02-01 13:31 nr_pdflush_threads -rw-r--r-- 1 root root 0 2008-02-01 13:31 overcommit_memory -rw-r--r-- 1 root root 0 2008-02-01 13:31 overcommit_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:31 page-cluster -rw-r--r-- 1 root root 0 2008-02-01 13:31 panic_on_oom -rw-r--r-- 1 root root 0 2008-02-01 13:31 percpu_pagelist_fraction -rw-r--r-- 1 root root 0 2008-02-01 13:31 swappiness -rw-r--r-- 1 root root 0 2008-02-01 13:31 swap_token_timeout -rw-r--r-- 1 root root 0 2008-02-01 13:31 vdso_enabled -rw-r--r-- 1 root root 0 2008-02-01 13:31 vfs_cache_pressure0 10 3000 40 500 0 0 0 256 256 32 65536 3831 2 0 50 3 0 0 60 300 1 100 Fast server: file value -rw-r--r-- 1 root root 0 2008-02-01 13:32 block_dump -rw-r--r-- 1 root root 0 2008-02-01 13:32 dirty_background_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:32 dirty_expire_centisecs -rw-r--r-- 1 root root 0 2008-02-01 13:32 dirty_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:32 dirty_writeback_centisecs -rw-r--r-- 1 root root 0 2008-02-01 13:32 laptop_mode -rw-r--r-- 1 root root 0 2008-02-01 13:32 lower_zone_protection -rw-r--r-- 1 root root 0 2008-02-01 13:32 max_map_count -rw-r--r-- 1 root root 0 2008-02-01 13:32 min_free_kbytes -r--r--r-- 1 root root 0 2008-02-01 13:32 nr_pdflush_threads -rw-r--r-- 1 root root 0 2008-02-01 13:32 overcommit_memory -rw-r--r-- 1 root root 0 2008-02-01 13:32 overcommit_ratio -rw-r--r-- 1 root root 0 2008-02-01 13:32 page-cluster -rw-r--r-- 1 root root 0 2008-02-01 13:32 swappiness -rw-r--r-- 1 root root 0 2008-02-01 13:32 vfs_cache_pressure 0 10 3000 40 500 0 0 65536 957 2 0 50 3 60 100 Scott Lovenberg escribió: Felipe Martinez Hermo wrote: Hi, everybody! I have been using samab on Debian for years and I have recently migrated my file server from version 3.0.14a-3sarge2 to 3.0.24-6etch4. One or our applications stores its data in a shared folder. This data is distributed in over 29000 files of about 1k-40k and is so much slower when it runs on the new server. I have thoroughly revised both smb.conf files, but can't see significant differences. I have read them so much that probably I'm already obfuscated. I have tuned socket options, but can't see any improvement. Any ideas? Thanks in advance how are your settings in /proc/sys/vm/*? If you've got the RAM, turn down the vfs_cache_pressure - you should get more hits. Also, what file system are you using? -- == Felipe Martínez Hermo [EMAIL PROTECTED] [EMAIL PROTECTED] == Servicios Informáticos UGT Galicia [EMAIL PROTECTED] [EMAIL PROTECTED] == -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
[Samba] Aplication slow after migration
Hi, everybody! I have been using samab on Debian for years and I have recently migrated my file server from version 3.0.14a-3sarge2 to 3.0.24-6etch4. One or our applications stores its data in a shared folder. This data is distributed in over 29000 files of about 1k-40k and is so much slower when it runs on the new server. I have thoroughly revised both smb.conf files, but can't see significant differences. I have read them so much that probably I'm already obfuscated. I have tuned socket options, but can't see any improvement. Any ideas? Thanks in advance -- == Felipe Martínez Hermo [EMAIL PROTECTED] [EMAIL PROTECTED] == Servicios Informáticos UGT Galicia [EMAIL PROTECTED] [EMAIL PROTECTED] == New server max. Version 3.0.24-6etch4. Old server clarence. Version 3.0.14a-3sarge2 WARNING: You have some share names that are longer than 12 characters. These may not be accessible to some older clients. (Eg. Windows9x, WindowsMe, and smbclient prior to Samba 3.0.) Server role: ROLE_DOMAIN_PDC Loaded services file OK. # Global parameters [global] workgroup = UGTGALICIA server string = Max. Servidor de disco de UGT Galicia obey pam restrictions = Yes passdb backend = tdbsam pam password change = Yes passwd program = /usr/bin/passwd %u passwd chat = *Enter\snew\sUNIX\spassword:* %n\n *Retype\snew\sUNIX\spassword:* %n\n *password\supdated\ssuccessfully* . unix password sync = Yes log level = 0 auth:2 syslog = 0 log file = /var/log/samba/log.%m max log size = 1000 time server = Yes deadtime = 15 fam change notify = No max disk size = 5000 socket options = IPTOS_LOWDELAY TCP_NODELAY SO_SNDBUF=4096 SO_RCVBUF=4096 add machine script = /usr/sbin/useradd -s /bin/false -d /var/lib/nobody %u logon script = scripts/%U.bat logon path = \\%h\profiles\%U logon drive = z: domain logons = Yes os level = 65 preferred master = Yes domain master = Yes dns proxy = No ldap ssl = no utmp = Yes panic action = /usr/share/samba/panic-action %d invalid users = root create mask = 0700 force create mode = 0700 directory mask = 0700 force directory mode = 0700 hosts allow = 172.15.1., 127.0.0.1 printing = cups print command = lpq command = %p lprm command = [homes] comment = Carpetas de Usuarios path = /home/%u valid users = %S read only = No browseable = No [informatica] comment = Servicios Informaticos path = /home/informatica valid users = @informatica force group = informatica read only = No create mask = 0770 force create mode = 0770 security mask = 0770 directory mask = 0770 force directory mode = 0770 directory security mask = 0770 Loaded services file OK. # Global parameters [global] workgroup = GALICIA server string = %h server (Samba %v) passdb backend = tdbsam, guest log file = /var/log/samba/log.%m max log size = 1000 socket options = IPTOS_LOWDELAY TCP_NODELAY SO_SNDBUF=4096 SO_RCVBUF=4096 printcap name = cups domain logons = Yes os level = 64 preferred master = Yes domain master = Yes dns proxy = No wins support = Yes ldap ssl = no panic action = /usr/share/samba/panic-action %d hosts allow = 172.15.1., 127. printing = cups print command = lpq command = lprm command = [homes] comment = Home Directories invalid users = root read only = No create mask = 0700 directory mask = 0700 force directory mode = 0700 browseable = No [informatica] comment = Servicios Informaticos path = /home/informatica force group = informatica read only = No create mask = 0770 force create mode = 0770 force security mode = 0770 force directory mode = 0770 directory security mask = 0770 force directory security mode = 0770 -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
[Samba] smbd process hangs up - XP login problems
Hi, everybody! I am having certain problems accessing my Samba domain. I have set up several XP SP2 clients to access a Samba 3.0.10 PDC running on a Debian testing server. Problem #1: Suddenly, the domain stopped working: it could not even be accessed browsing the network. After some investigation, I found that nobody could connect to the shares or log on to the domain, except one of the clients. I closed session on that client and stopped the service "/etc/init.d/samba stop", but the smbd process kept on running, so I had to kill it. Everybody could log in after the service was restarted. Problem #2: Once the services were started again, everything works fine, except for two clients: at the first prompt for login it replies "Windows can't log on to the domain because the domain controller is blocked or not available or because the machine account could not be found. Try again later" (Translated from spanish). If I log on as machine administrator and browse the domain server through "my network places", close session and try to log on to the domain again it works! Are these two problems related to each other? Is problem #1 a bug? How can I get to log in at the first try? Thank you for reading this. -- == Felipe Martínez Hermo [EMAIL PROTECTED] [EMAIL PROTECTED] == Servicios Informáticos UGT Galicia [EMAIL PROTECTED] == -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
