[solved] Litespeed webserver stops working while cleancache script is runnning

[NiteWave]

the main problem is i/o. average iowait is 22.05, this is very high.

~>date;uptime;iostat
Fri Jun 15 22:47:01 EDT 2012
 22:47:01 up 24 days, 15:32,  1 user,  load average: 1.26, 0.97, 0.87
Linux 2.6.18-338.12.1.el5.lve0.8.34ent (xxxx)    06/15/12

avg-cpu:  %user   %nice %system %iowait  %steal   %idle
          12.92    0.29    4.64   22.05    0.00   60.09

Device:            tps   Blk_read/s   Blk_wrtn/s   Blk_read   Blk_wrtn
sda             247.39       867.95        27.06 1848352732   57620252
sda1              0.00         0.05         0.01     106678      11988
sda2             37.17       136.88       527.71  291502237 1123779840
sda3              9.93       171.04       105.91  364241245  225538992
sda4              0.00         0.00         0.00          6          0
sda5              1.47        21.51        20.47   45811450   43586016
sda6              0.00         0.00         0.00       2351       1688
sda7             10.13        10.10       530.31   21511492 1129316048
sda8            188.68       528.36       859.50 1125171313 1830352976

 

[gnu.smtp]

Hello,

We have this issue only when the script runs. Does this script need very high number of IO opearations?

Also what is a possible solution to this..Does increasing the frequency of running the script help?

 

[NiteWave]

please try:

#cd /usr/local/lsws/admin/misc/
#cp -a cleancache.sh cleancache_000.sh

then edit cleancache.sh, make it do nothing.

then you create a cronjob, to run cleancache_000.sh, at an interval you wish( once per hour or once per day etc), or do not run it at all.

so the impact from the clean script will be 0. monitor for one or 2 days, see if your server are all ok since.

 

[gnu.smtp]

Hello,

We have tried the same and seems to be much better.

But the question is why the original script is having issue for this server only. We have more than 10 servers running on litespeed and none of those server is having such issues.

Really weird Hope you can make some thing to make the script running more efficiently in the future upgrades.

 

[NiteWave]

glad to hear the progress.

I think it may only happen on high iowait server, which already been a big problem.

to improve at the lsws side, maybe you can put
`date` >> /root/cleanup.log

at the beginning of cleancache.sh, to see if it's run too frequently.