I have a USB disk connected to a Poweredge server. The USB Disk is a Mad Dog MegaVault with a 160GB drive installed. It's all working fine, but the external LED on the unit is not working.
I'm currently running a script (yum-pull) for setting up a yum mirror and it sits for long periods of time seemingly doing nothing. I started the script this morning at 6:20 for my FC4 yum mirror and it's now 8:50 and it's still running. A "top" shows very little activity from a CPU perspective.
The guy who wrote the script says there's probably tons of disk activity going on. The script does continue to run because it will eventually finish. One might think the slowness is because it's a USB drive, but I can copy large amounts of data and do permission changes on files, at the same time this script is running, and it all happens quickly.
Guess I'm just trying to verify whether there really is a lot of disk activity going on. I tried looking at ksysguard, but wasn't very successful at seeing what I wanted.
I'm not sure I want to use yum-pull if it take 3 hours to run per mirror. Right now I'm mirroring FC4, but I also want to do Centos 4.0 and 4.2, and maybe even Suse 9.2.
Are there any other disk activity monitors I should try?
Thanks, James
On 13/02/06, James Pifer jep@obrien-pifer.com wrote:
Are there any other disk activity monitors I should try?
iostat ?
James Pifer wrote:
I have a USB disk connected to a Poweredge server. The USB Disk is a Mad Dog MegaVault with a 160GB drive installed. It's all working fine, but the external LED on the unit is not working.
I'm currently running a script (yum-pull) for setting up a yum mirror and it sits for long periods of time seemingly doing nothing. I started the script this morning at 6:20 for my FC4 yum mirror and it's now 8:50 and it's still running. A "top" shows very little activity from a CPU perspective.
The guy who wrote the script says there's probably tons of disk activity going on. The script does continue to run because it will eventually finish. One might think the slowness is because it's a USB drive, but I can copy large amounts of data and do permission changes on files, at the same time this script is running, and it all happens quickly.
Guess I'm just trying to verify whether there really is a lot of disk activity going on. I tried looking at ksysguard, but wasn't very successful at seeing what I wanted.
I'm not sure I want to use yum-pull if it take 3 hours to run per mirror. Right now I'm mirroring FC4, but I also want to do Centos 4.0 and 4.2, and maybe even Suse 9.2.
Are there any other disk activity monitors I should try?
Thanks, James
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
If you're looking for a real-time graphical display, try GKrellm configured to show your USB drive's partition(s). This works fine in KDE as well as Gnome, BTW.
If you're looking for a real-time graphical display, try GKrellm configured to show your USB drive's partition(s). This works fine in KDE as well as Gnome, BTW.
Is there a gkrellm rpm for CentOS 4? Can't seem to find one. Or do I need to get the source/compile/etc?
I tried iostat and I guess there is disk activity, but I have no idea if it's considered a lot or not, see below.
Thanks, James
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 2.63 35.72 19.24 134576004 72488792
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 15.50 1976.00 0.00 3952 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 28.36 1954.23 23.88 3928 48
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 11.50 2144.00 4.00 4288 8
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 9.50 2176.00 0.00 4352 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 12.50 1952.00 16.00 3904 32
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 12.00 2080.00 0.00 4160 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 9.00 2176.00 0.00 4352 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 17.00 1920.00 56.00 3840 112
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 10.05 2194.97 0.00 4368 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 22.00 1956.00 20.00 3912 40
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 13.93 1898.51 0.00 3816 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 13.00 2072.00 0.00 4144 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 15.50 2164.00 24.00 4328 48
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 18.50 1988.00 0.00 3976 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 11.94 1954.23 0.00 3928 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 20.40 2089.55 23.88 4200 48
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 15.50 1928.00 0.00 3856 0
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 16.58 2142.71 28.14 4264 56
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sdb1 15.00 2020.00 0.00 4040 0
James Pifer wrote:
If you're looking for a real-time graphical display, try GKrellm configured to show your USB drive's partition(s). This works fine in KDE as well as Gnome, BTW.
Is there a gkrellm rpm for CentOS 4? Can't seem to find one. Or do I need to get the source/compile/etc?
I thought about that right after sending the 1st reply. I got gkrellm from DAG... Specifically: http://dag.wieers.com/packages/gkrellm/
There are RHEL4 packages for both i386 and x86_64. Hope this helps.
I thought about that right after sending the 1st reply. I got gkrellm from DAG... Specifically: http://dag.wieers.com/packages/gkrellm/
There are RHEL4 packages for both i386 and x86_64. Hope this helps.
gkrellm is pretty cool. Thanks.
James
On 2/13/06, James Pifer jep@obrien-pifer.com wrote:
I thought about that right after sending the 1st reply. I got gkrellm from DAG... Specifically: http://dag.wieers.com/packages/gkrellm/
There are RHEL4 packages for both i386 and x86_64. Hope this helps.
gkrellm is pretty cool. Thanks.
Yep, pretty slick, but rather tiny for my week old eyes. Does anyone know a way to expand the size of gkrellm, ie I'm looking for a poor-man's RTFM <grin>.
-- Collins Richey The agnostic dyslexic insomniac lies awake wondering if there is a dog.
On Monday 13 February 2006 21:23, Collins Richey wrote:
On 2/13/06, James Pifer jep@obrien-pifer.com wrote:
I thought about that right after sending the 1st reply. I got gkrellm from DAG... Specifically: http://dag.wieers.com/packages/gkrellm/
There are RHEL4 packages for both i386 and x86_64. Hope this helps.
gkrellm is pretty cool. Thanks.
Yep, pretty slick, but rather tiny for my week old eyes. Does anyone know a way to expand the size of gkrellm, ie I'm looking for a poor-man's RTFM <grin>.
Right click on the top and is should open a configuration window. I found a help once that told me about it, but don't remember where. I made it MUCH bigger! :)
On Mon, 2006-02-13 at 22:15, Robert Tate wrote:
gkrellm is pretty cool. Thanks.
Yep, pretty slick, but rather tiny for my week old eyes. Does anyone know a way to expand the size of gkrellm, ie I'm looking for a poor-man's RTFM <grin>.
Right click on the top and is should open a configuration window. I found a help once that told me about it, but don't remember where. I made it MUCH bigger! :)
It should tell you when you start it that you right-mouse on the top bar or hit F1 with the mouse over the window. Likewise, right-mouse over individual elements to configure and set options.
On Mon, 13 Feb 2006, James Pifer wrote:
I have a USB disk connected to a Poweredge server. The USB Disk is a Mad Dog MegaVault with a 160GB drive installed. It's all working fine, but the external LED on the unit is not working.
I'm currently running a script (yum-pull) for setting up a yum mirror and it sits for long periods of time seemingly doing nothing. I started the script this morning at 6:20 for my FC4 yum mirror and it's now 8:50 and it's still running. A "top" shows very little activity from a CPU perspective.
The guy who wrote the script says there's probably tons of disk activity going on. The script does continue to run because it will eventually finish. One might think the slowness is because it's a USB drive, but I can copy large amounts of data and do permission changes on files, at the same time this script is running, and it all happens quickly.
Guess I'm just trying to verify whether there really is a lot of disk activity going on. I tried looking at ksysguard, but wasn't very successful at seeing what I wanted.
I'm not sure I want to use yum-pull if it take 3 hours to run per mirror. Right now I'm mirroring FC4, but I also want to do Centos 4.0 and 4.2, and maybe even Suse 9.2.
Are there any other disk activity monitors I should try?
You may want to look at dstat. With:
dstat -cd -f
You can compare the output from your CPU (check iowait) and the disks.
Also, beware that the disk i/o is relative to what your disk can handle, so there is not clear upper boundary. Graphs may trick you into showing how much free disk bandwidth there is, but that is deceptive.
With:
hdparm -tT
you can check to see what your disk is capable of. And that might help interpreting raw values much better. (together with the iowait values)
Kind regards, -- dag wieers, dag@wieers.com, http://dag.wieers.com/ -- [all I want is a warm bed and a kind word and unlimited power]
-
Collins Richey -
Dag Wieers -
James Pifer -
Les Mikesell -
Robert -
Robert Tate -
Will McDonald