RE: [opensuse] Running slow, / full

2008-01-28 Thread Nick Zeljkovic
find / -size 50k

you can specify where to find however, find /home ... would find all files 
larger than 500mb in /home.

--
Best regards,
Nick Zeljkovic


-Original Message-
From: Bob [mailto:[EMAIL PROTECTED] 
Sent: Monday, January 28, 2008 7:21 PM
To: opensuse@opensuse.org
Subject: Re: [opensuse] Running slow, / full

On Monday 28 January 2008 17:59:44 Bill Anderson wrote:
 Sunny wrote:
  On Jan 28, 2008 10:15 AM, Bob [EMAIL PROTECTED] wrote:
  My system has started running very slow, and the root directory is
  reported as having no free space. I've found a file /proc/kcore which is
  the probable culprit, being about 1000MB, with today's date, but even
  root is not allowed to move or delete this file.
 
  My system looks like this:
 
  Device  SizeMount point Free
  sda2965.1MB /   0B
  sda360.8MB  /boot   46.8MB
  sda59.8GB   /usr4.9GB
  sda62.0GB   /var1.2GB
  sda72.0GB   /opt906.3MB
  sda81011.4MB/tmp926.4MB
  sda9257.3GB /home   154.1GB
 
  /proc is virtual file system, which resides in memory, not on disk.
  Even some files are not actually in memory, but when you read them,
  you actually receive information from the kernel.
  In your case, read here for /proc/kcore:
  http://www.unixguide.net/linux/faq/04.16.shtml
 
  Run the command df to see the how your partitions are filled in.
  Run top to see which process consumes more resources.
 
  Note, that if you run firefox for several days w/o stopping it, it may
  eat a lot of resources.
 
  Cheers

 I prefer the term pseudo filesystem, since /proc does not reside in
 memory. As with any file system, procfs implements the functions defined
 by vfs, the virtual filesystem. The functions implemented actually read
 from, and in some cases write to, kernel data structures. The pathnames
 under /proc define which functions to call. There are a large number of
 such file systems: rootfs, sysfs, relayfs, tmpfs, and the list goes on.
 It works, because every filesystem is an implementation of vfs.

 Bill Anderson
 WW7BA

Hmmm. I can't say I understand much of that :(.

I've been informed off-list that /proc/kcore is simply my running kernel, so 
that's probably not the file I'm looking for.

Can anyone suggest a simple CLI incantation to find all files larger than, 
say, 500MB?

-- 
Bob

openSUSE 10.3, Kernel 2.6.22.13-0.3-default, KDE 3.5.8
Intel Celeron 2.53GB, 2GB DDR RAM, nVidia GeForce 7600GS
-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



RE: [opensuse] AHCI / NCQ problem

2008-01-22 Thread Nick Zeljkovic
Thanks Greg.

I've been asking a bunch of people and no one had a clue what was going on.
I wanted to enable this on a brand new MySQL server to help cope with I/O
load but I'm not in a hurry, we'll see if Novell releases a patch, if not I
can put a newer kernel once it arrives. Just out of curiosity, were the
steps I took correct ?

--
Best regards,
Nick Zeljkovic

-Original Message-
From: Greg Freemyer [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, January 22, 2008 6:06 PM
To: opensuse@opensuse.org
Subject: Re: [opensuse] AHCI / NCQ problem

On Jan 19, 2008 1:08 PM, Nick Zeljkovic [EMAIL PROTECTED] wrote:
 Hi,

 I'm having problems activating NCQ on Asus P5B-MX , Seagate Barracuda
 ST3250410AS (250GB) with Opensuse 10.3 installed.
 Before installation I've set in BIOS Enhanced IDE configuration for S-ATA
 mode (from my understanding that enables advanced features like NCQ),
 installed the system no problem. After first boot I've noticed ahci driver
 wasn't loaded, only ata_piix, so I've edited /etc/sysconfig/kernel and
added
 ahci before ata_piix driver, rebuilt initrd, edited modprobe.conf and
 replaced ata_piix occurrences with ahci.
 Original line:
 # ata_piix can't handle ICH6 in AHCI mode
 install ata_piix /sbin/modprobe ahci 21 |:; /sbin/modprobe
 --ignore-install ata_piix

 My line:
 # ata_piix can't handle ICH6 in AHCI mode
 install ahci /sbin/modprobe ahci 21 |:; /sbin/modprobe --ignore-install
 ahci

 After reboot I saw ahci loaded under libata but after ata_piix
 libata | ata_piix, ahci

 and actually used was ata_piix despite being loaded first. I've then:
 # dmesg|grep NCQ

 And got a line back like:

 ata3.00: ATA-7, 321672960 sectors: LBA48 NCQ (depth 0/32)

 # echo 31  /sys/block/sda/device/queue_depth

 resulted in permission denied. (as I found on linux-ata FAQ) Further
reading
 on linux-ata FAQ says:
 If the line containing the sector count and maximum UDMA speed does not
 mention NCQ, your drive does not support it.
 I'm puzzled since I have UDMA showing on:
 # dmesg|grep ata3

 Which displays all disk info, and NCQ (depth 0/32) would mean NCQ is
 present but disabled.

 From what I can see disk should support NCQ and I was informed by person
who
 sold it to me that it was NCQ ready. (I specifically wanted NCQ disk).

 I don't know if I'm making a mistake somewhere or disk just doesn't
support
 it. I'm in for some help please!

 P.S. Yes I've rebooted ;)

 --
 Best regards,
 Nick Zeljkovic

Nick,

Linux'es implementation of NCQ had some fundamental problems that have
been fixed in the soon to be released 2.6.24

Due to the bug, a bunch of drives got blacklisted.  The majority of
that blacklist has been removed from the 2.6.24 rc series.

I have not read about those patches having been released by Novell for
the SUSE kernels.  (A Novell employee found the problem and wrote the
fix so I would hope his work will get into their kernel eventually.)

So, you can either wait for that to happen (if it happens), or you can
move to a recent KOTD kernel.

FYI: I don't know how stable the KOTD kernels are right now.  2.6.24
has been very slow to get released, so I assume there must be some
regression(s) they are trying to resolve currently.

Greg
-- 
Greg Freemyer
Litigation Triage Solutions Specialist
http://www.linkedin.com/in/gregfreemyer
First 99 Days Litigation White Paper -
http://www.norcrossgroup.com/forms/whitepapers/99%20Days%20whitepaper.pdf

The Norcross Group
The Intersection of Evidence  Technology
http://www.norcrossgroup.com
-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



RE: [opensuse] NAT under 10.1

2008-01-22 Thread Nick Zeljkovic
You're welcome:)
Please, forward this solution to public mail-list. I was tried to do it, but
with no result (mail delivery error).
Thanks!
On Jan 23, 2008 12:08 AM, Nick Zeljkovic  [EMAIL PROTECTED] wrote:
Excellent, that did it! Thank you very much.
 
--
Best regards,
Nick Zeljkovic 
 
From: Alexander R [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, January 22, 2008 9:55 PM

To: Nick Zeljkovic
Subject: Re: [opensuse] NAT under 10.1
 
Add this line:

iptables -t filter -A FORWARD -i eth0 -o eth1 -m state --state
ESTABLISHED,RELATED -j ACCEPT
On Jan 22, 2008 11:42 PM, Nick Zeljkovic  [EMAIL PROTECTED] wrote:
Yeah, that one, forgot to put  when I was writing this.
 
# cat /proc/sys/net/ipv4/ip_forward
1
 
--
Best regards,
Nick Zeljkovic
 
From: Alexander R [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, January 22, 2008 9:13 PM
To: Nick Zeljkovic
Subject: Re: [opensuse] NAT under 10.1
 
echo 1  /proc/sys/net/ipv4/ip_forward ?
On Jan 22, 2008 12:23 AM, Nick Zeljkovic [EMAIL PROTECTED] wrote:
Greets:

I have two SuSE boxes, one 10.1 (with internet connection) and one 10.3
connected to a second private NIC (eth1).  I've done the following:
#echo 1 /proc/sys/net/ipv4/ip_forward
# iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE 
# iptables -A FORWARD -i eth0 -o eth1 -m state --state RELATED,ESTABLISHED
-j ACCEPT
# iptables -A FORWARD -i eth1 -o eth0 -j ACCEPT

And on client machine added default GW. Eth0 would be external interface and

eth1 private.

It looks like an issue with firewall not allowing or forwarding packets
correctly, doing a:
# tcpdump -i eth1 icmp

While running a ping from client machine and it was received properly. 

--
Best regards,
Nick Zeljkovic

-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[opensuse] NAT under 10.1

2008-01-21 Thread Nick Zeljkovic
Greets:

I have two SuSE boxes, one 10.1 (with internet connection) and one 10.3
connected to a second private NIC (eth1).  I've done the following:
#echo 1 /proc/sys/net/ipv4/ip_forward
# iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE
# iptables -A FORWARD -i eth0 -o eth1 -m state --state RELATED,ESTABLISHED
-j ACCEPT
# iptables -A FORWARD -i eth1 -o eth0 -j ACCEPT

And on client machine added default GW. Eth0 would be external interface and
eth1 private.

It looks like an issue with firewall not allowing or forwarding packets
correctly, doing a:
# tcpdump -i eth1 icmp

While running a ping from client machine and it was received properly.

--
Best regards,
Nick Zeljkovic


-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[opensuse] AHCI / NCQ problem

2008-01-19 Thread Nick Zeljkovic
Hi,

I’m having problems activating NCQ on Asus P5B-MX , Seagate Barracuda
ST3250410AS (250GB) with Opensuse 10.3 installed.
Before installation I’ve set in BIOS Enhanced IDE configuration for S-ATA
mode (from my understanding that enables advanced features like NCQ),
installed the system no problem. After first boot I’ve noticed ahci driver
wasn’t loaded, only ata_piix, so I’ve edited /etc/sysconfig/kernel and added
ahci before ata_piix driver, rebuilt initrd, edited modprobe.conf and
replaced ata_piix occurrences with ahci. 
Original line:
# ata_piix can't handle ICH6 in AHCI mode
install ata_piix /sbin/modprobe ahci 21 |:; /sbin/modprobe
--ignore-install ata_piix

My line:
# ata_piix can't handle ICH6 in AHCI mode
install ahci /sbin/modprobe ahci 21 |:; /sbin/modprobe --ignore-install
ahci

After reboot I saw ahci loaded under libata but after ata_piix
libata | ata_piix, ahci

and actually used was ata_piix despite being loaded first. I’ve then:
# dmesg|grep NCQ

And got a line back like:

ata3.00: ATA-7, 321672960 sectors: LBA48 NCQ (depth 0/32)

# echo 31  /sys/block/sda/device/queue_depth

resulted in permission denied. (as I found on linux-ata FAQ) Further reading
on linux-ata FAQ says:
“If the line containing the sector count and maximum UDMA speed does not
mention NCQ, your drive does not support it.”
I’m puzzled since I have UDMA showing on:
# dmesg|grep ata3

Which displays all disk info, and “NCQ (depth 0/32)” would mean NCQ is
present but disabled.
 
From what I can see disk should support NCQ and I was informed by person who
sold it to me that it was NCQ ready. (I specifically wanted NCQ disk).

I don’t know if I’m making a mistake somewhere or disk just doesn’t support
it. I’m in for some help please!

P.S. Yes I’ve rebooted ;)

--
Best regards,
Nick Zeljkovic 


--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]