hobbit_rrd stops working after about 1 hour
Well, as nobody has suggested anything to my problem I guess that I'm the only one having this issue. I have managed to find the root cause. The hobbitd_rrd process was showing to be in "uninterruptible sleep" state most of the time with high iowait associated with the CPU it was running on. I suspected that the problem may be due to disk IO while updating rrds for the 2000 hosts. I created a tmpfs filesystem and copied the rrd directory into it. Since then (48 hours ago) my rrd graphs have been updating continuously. I do however need to write back to disk periodically to avoid loss of data after a reboot.
This is OK as a temporary fix but I would like to have a permanent solution. I would like to hear from other hobbit users who have more than 1000 hosts monitored. What type of servers and disk subsystems are they using? Perhaps my problem is to do with RedHat and Dell server combination. Perhaps I need to stripe over multiple spindles.
-Naeem
Naeem
Maqsud/SYBASE
To
08/18/2005 05:02 hobbit at hswn.dk
PM cc
Subject
hobbit_rrd stops working after
about 1 hour
Hi,
I'm testing out hobbit 4.1.1 for possible migration from big brother (with bbgen). I suspected scalability issues with BB as my rrd graphs were updated intermittently. However, hobbit is exhibiting similar problems. After about 1 hr of restarting hobbit, the rrd graphs stop updating except for the cpu utilization for the hobbit server itself.
The hobbit server is running RedHat Linux AS 3.0. It has 2 x 2.4 GHz Xeon processors and 1GB of memory. About 800 servers are sending updates to the hobbit server. Another 1200 servers are getting remote tests.
Load average has stayed below 1 most of the time. CPU usage has been low with 75% idle. 4 CPUs show up due to hyperthreading and I've noticed that after the restart of hobbit server, hobbitd_rrd process stays on CPU3 with 100% utilization for the one hour that it is busy.
I hope someone can shed some light on this.
Thanks, Naeem
Hi Naeem,
I have over 18000 rrd files being updated every 5 minutes, and havent
seen any problems with them.
i'm running hobbit on a 2x3Gh compaq server with redhat 3.0
but,
i do have heavy i/o due to hobbitd_rrd, and it is getting a problem
for me,
i'm planning to add a array card with 256M of cache in 1 or 2 days to
lower the i/o wait..
i have the feeling that hobbitd_rrd could cause performance issue for
large site and may not be fully optimized... henrik ?
concerning your problem, i posted this early this month :
"hobbitd just slows down dramaticly, causing bbtest's results
transmition to take over 250s instead of 20s;
the rrd files aren't being updated anymore and some requests to cgi's
are saying event is not available..
notifications are being sent though and external scripts don't seem
to be affected
doing a stop/start of hobbit solved the problem right now."
this happened twice for me; bbtest went yellow, i got called and
restarted hobbit..
is everything nice and green for your bigbrother server itself
(bbtest,bbgen,hobbitd) ?
have their timing execution really changed before and after the
problem ?
do you have any interesting logs ?
are the graphs for the bigbrother server itselft with "holes" ? (or
the first server in your bb-hosts file)
-- Olivier Beau
Le 22 août 05 à 21:28, Naeem.Maqsud at sybase.com a écrit :
Well, as nobody has suggested anything to my problem I guess that
I'm the only one having this issue. I have managed to find the root cause. The hobbitd_rrd process was showing to be in "uninterruptible sleep"
state most of the time with high iowait associated with the CPU it was running
on. I suspected that the problem may be due to disk IO while updating
rrds for the 2000 hosts. I created a tmpfs filesystem and copied the rrd directory into it.
Since then (48 hours ago) my rrd graphs have been updating continuously.
I do however need to write back to disk periodically to avoid loss of
data after a reboot.This is OK as a temporary fix but I would like to have a permanent solution. I would like to hear from other hobbit users who have
more than 1000 hosts monitored. What type of servers and disk subsystems are
they using? Perhaps my problem is to do with RedHat and Dell server
combination. Perhaps I need to stripe over multiple spindles.-Naeem
Naeem Maqsud/SYBASETo 08/18/2005 05:02 hobbit at hswn.dk PM cc
Subject hobbit_rrd stops working after about 1 hour
Hi,
I'm testing out hobbit 4.1.1 for possible migration from big
brother (with bbgen). I suspected scalability issues with BB as my rrd graphs were updated intermittently. However, hobbit is exhibiting similar
problems. After about 1 hr of restarting hobbit, the rrd graphs stop updating
except for the cpu utilization for the hobbit server itself.The hobbit server is running RedHat Linux AS 3.0. It has 2 x 2.4
GHz Xeon processors and 1GB of memory. About 800 servers are sending updates
to the hobbit server. Another 1200 servers are getting remote tests.Load average has stayed below 1 most of the time. CPU usage has
been low with 75% idle. 4 CPUs show up due to hyperthreading and I've
noticed that after the restart of hobbit server, hobbitd_rrd process stays on
CPU3 with 100% utilization for the one hour that it is busy.I hope someone can shed some light on this.
Thanks, Naeem
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
Naeem.Maqsud wrote:
hobbitd_rrd process was showing to be in "uninterruptible sleep" state most of the time with high iowait associated with the CPU it was running on. I suspected that the problem may be due to disk IO while updating rrds for the 2000 hosts. I created a tmpfs filesystem and copied the rrd directory into it.
You might want to look at your disk hardware and the software setup. I have 2000 hosts myself, with a total of just over 18000 RRD-files that are updated every 5 minutes. vmstat tells me this system spends about 15% of its time in I/O wait.
This is a Debian/Linux system on Sun hardware - two SCSI disks in a raid-1 config (Linux software raid mirror) with a reiserfs filesystem.
On Mon, Aug 22, 2005 at 10:30:52PM +0200, Olivier Beau wrote:
i have the feeling that hobbitd_rrd could cause performance issue for
large site and may not be fully optimized... henrik ?
It might become a problem - I agree with that.
The solutions are probably going to be those that you would with any kind of application that has a high I/O load. E.g. mail- and news-servers face similar problems. So your choice of filesystem and mount-options become important.
For Linux systems you'd definitely want to use one of the better performing filesystems, e.g. Reiserfs or JFS. ext2/3 - in my experience, there are tons of benchmarks pointing in whatever direction you like - is slower. Using the "noatime,nodiratime" mount options is also recommended, as is the reiserfs "notail" option.
For Solaris ufs filesystems, I am told the "journal" option will boost performance significantly, although I have never tried it myself.
Since all of the hobbitd_rrd disk activity is done by the rrdtool library there's not a whole lot the Hobbit can do to boost throughput - at least not as long as we stick to with rrdtool as the back-end for graphs. And I have no intention of changing that.
Regards, Henrik
Hi,
it happened a third time for me this night (3 times in 3 weeks) : syptoms: hobbitd seems to slow down and stops graphing.
i think Naeem and me are hitting a bug.
i've looked closer this night, and i saw that hobbitd_rrd was running at 100% on the cpu it was on; i tried to strace the procees, but strace wouldnt give me any ouptut ! i finally killed hobbitd_rrd, and everything went back to normal. hobbitd.log has : Task rrdstatus terminated, status 1 rrd_status.log has : Worker process died with exit code 1, terminating
during normal running, vmstat shows a i/o wait of 25% my problems happened always at night, exactly at the time legato starts
-> something strange is happening whith hobbitd_rrd when the server is under very heavy i/o..
henrik, could this be a OS issue or more a hobbitd_rrd problem ?
Olivier
Selon Naeem.Maqsud at sybase.com:
Well, as nobody has suggested anything to my problem I guess that I'm the only one having this issue. I have managed to find the root cause. The hobbitd_rrd process was showing to be in "uninterruptible sleep" state most of the time with high iowait associated with the CPU it was running on. I suspected that the problem may be due to disk IO while updating rrds for the 2000 hosts. I created a tmpfs filesystem and copied the rrd directory into it. Since then (48 hours ago) my rrd graphs have been updating continuously. I do however need to write back to disk periodically to avoid loss of data after a reboot.
This is OK as a temporary fix but I would like to have a permanent solution. I would like to hear from other hobbit users who have more than 1000 hosts monitored. What type of servers and disk subsystems are they using? Perhaps my problem is to do with RedHat and Dell server combination. Perhaps I need to stripe over multiple spindles.
-Naeem
Naeem Maqsud/SYBASE To 08/18/2005 05:02 hobbit at hswn.dk PM cc Subject hobbit_rrd stops working after about 1 hourHi,
I'm testing out hobbit 4.1.1 for possible migration from big brother (with bbgen). I suspected scalability issues with BB as my rrd graphs were updated intermittently. However, hobbit is exhibiting similar problems. After about 1 hr of restarting hobbit, the rrd graphs stop updating except for the cpu utilization for the hobbit server itself.
The hobbit server is running RedHat Linux AS 3.0. It has 2 x 2.4 GHz Xeon processors and 1GB of memory. About 800 servers are sending updates to the hobbit server. Another 1200 servers are getting remote tests.
Load average has stayed below 1 most of the time. CPU usage has been low with 75% idle. 4 CPUs show up due to hyperthreading and I've noticed that after the restart of hobbit server, hobbitd_rrd process stays on CPU3 with 100% utilization for the one hour that it is busy.
I hope someone can shed some light on this.
Thanks, Naeem
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
-- Olivier Beau
On Tue, Aug 23, 2005 at 08:28:46AM +0200, Olivier Beau wrote:
i've looked closer this night, and i saw that hobbitd_rrd was running at 100% on the cpu it was on; i tried to strace the procees, but strace wouldnt give me any ouptut ! i finally killed hobbitd_rrd, and everything went back to normal.
Your description sounds as if hobbitd_rrd goes into some loop. That would explain why strace shows nothing (strace only shows system call activity - if an application is looping in user-mode it won't show any system call activity).
The next time it happens, could you kill it with "kill -ABRT" ? That should cause it to dump core, and that might give a clue as to where it is looping.
henrik, could this be a OS issue or more a hobbitd_rrd problem ?
hobbitd_rrd, I'm afraid.
Regards, Henrik
Olivier,
Why don't you try the approach of putting your rrd files in a tmpfs filesystem? This seems to have resolved my rrd problem. At least you can try to see if this resolves your issue and then you know for sure it is related to disk IO. This is what I did:
mkdir /usr/local/bbvar/rrd_orig
mv /usr/local/bbvar/rrd /usr/local/bbvar/rrd_orig
mkdir /usr/local/bbvar/rrd
Add the following line to /etc/fstab: tmpfs /usr/local/bbvar/rrd tmpfs mode=755,rw,size=2G 0 0
mount /usr/local/bbvar/rrd; chown <id of bb user> /usr/local/bbvar/rrd
cp -pr /usr/local/bbvar/rrd_orig/rrd/* /usr/local/bbvar/rrd
Start hobbit
If you want to keep this as a permanent solution, then you will need to setup a cronjob to periodically copy the rrd files from the tmpfs filesystem back to disk. This is because if you unmount the tmpfs FS all data will be lost. You can put a line in crontab as shown below to run at 8:30 PM daily:
30 20 * * * rsync -av /usr/local/bbvar/rrd /usr/local/bbvar/rrd_orig
Remember that everytime you reboot, you will need to copy the files from disk to the tmpfs filesystem. You can put a line in /etc/rc.local to do this for you.
Hope this helps.
-Naeem
Olivier Beau
<olivier at qalpit.c
om> To
hobbit at hswn.dk
08/22/2005 11:28 cc
PM
Subject
Re: [hobbit] Re: hobbit_rrd stops
Please respond to working after about 1 hour
hobbit at hswn.dk
Hi,
it happened a third time for me this night (3 times in 3 weeks) : syptoms: hobbitd seems to slow down and stops graphing.
i think Naeem and me are hitting a bug.
i've looked closer this night, and i saw that hobbitd_rrd was running at 100% on the cpu it was on; i tried to strace the procees, but strace wouldnt give me any ouptut ! i finally killed hobbitd_rrd, and everything went back to normal. hobbitd.log has : Task rrdstatus terminated, status 1 rrd_status.log has : Worker process died with exit code 1, terminating
during normal running, vmstat shows a i/o wait of 25% my problems happened always at night, exactly at the time legato starts
-> something strange is happening whith hobbitd_rrd when the server is under very heavy i/o..
henrik, could this be a OS issue or more a hobbitd_rrd problem ?
Olivier
Selon Naeem.Maqsud at sybase.com:
Well, as nobody has suggested anything to my problem I guess that I'm the only one having this issue. I have managed to find the root cause. The hobbitd_rrd process was showing to be in "uninterruptible sleep" state most of the time with high iowait associated with the CPU it was running on. I suspected that the problem may be due to disk IO while updating rrds for the 2000 hosts. I created a tmpfs filesystem and copied the rrd directory into it. Since then (48 hours ago) my rrd graphs have been updating continuously. I do however need to write back to disk periodically to avoid loss of data after a reboot.
This is OK as a temporary fix but I would like to have a permanent solution. I would like to hear from other hobbit users who have more than 1000 hosts monitored. What type of servers and disk subsystems are they using? Perhaps my problem is to do with RedHat and Dell server combination. Perhaps I need to stripe over multiple spindles.
-Naeem
Naeem
Maqsud/SYBASE
To
08/18/2005 05:02 hobbit at hswn.dk
PM
cc
Subject
hobbit_rrd stops working after
about 1 hour
Hi,
I'm testing out hobbit 4.1.1 for possible migration from big brother
(with
bbgen). I suspected scalability issues with BB as my rrd graphs were updated intermittently. However, hobbit is exhibiting similar problems. After about 1 hr of restarting hobbit, the rrd graphs stop updating except for the cpu utilization for the hobbit server itself.
The hobbit server is running RedHat Linux AS 3.0. It has 2 x 2.4 GHz Xeon processors and 1GB of memory. About 800 servers are sending updates to the hobbit server. Another 1200 servers are getting remote tests.
Load average has stayed below 1 most of the time. CPU usage has been low with 75% idle. 4 CPUs show up due to hyperthreading and I've noticed that after the restart of hobbit server, hobbitd_rrd process stays on CPU3 with 100% utilization for the one hour that it is busy.
I hope someone can shed some light on this.
Thanks, Naeem
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
-- Olivier Beau
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
participants (3)
-
henrik@hswn.dk
-
Naeem.Maqsud@sybase.com
-
olivier@qalpit.com