11.52 exim/spamd/nscd
I'm on the RELEASE tier. Since the update to 11.52 a few weeks ago, I've periodically been receiving alert messages that the spamd service is unavailable and restart attempts have failed. So I have to SSH in and issue a 'service exim restart' to fix it.
ps says the spamd process is running, however I do see errors in exim_paniclog about socket timeouts. It also appears that exim isn't delivering mail even though there's activity in exim_mainlog (mostly incoming garbage from spammers). My spamd startup config is 2 idle, 12 max and 25 connections per child. This worked beautifully on 11.50 with all the junk we get.
Now today I get an alert that spamd failed and restart failed. And then 10-15 minutes later, nscd has failed and couldn't be restarted. dmesg says nscd segfaulted. ps says both services are running, however I had to manually issue a restart on both before checkservd reported them up.
My cPanel server is on CentOS 6 64-bit bare metal with dual 6-core HT Xeons and 24GB of RAM, so there's plenty horsepower. Everything was happy on 11.50, so I'm thinking the introduction of nscd with 11.52 is the issue here.
I know that setuids are issued for spamd, so is this possibly a loading issue on nscd? Can and should I disable nscd?
-
OK, now I see that I can disable NSCD in WHM>Service Manager. Did that. Hopefully it's not enabled again when upcp runs next time. And hopefully that will stop the exim and spamd issues. 0 -
New I'm on the RELEASE tier. Since the update to 11.52 a few weeks ago, I've periodically been receiving alert messages that the spamd service is unavailable and restart attempts have failed. So I have to SSH in and issue a 'service exim restart' to fix it.
Hello :) Please also note the following thread regarding SpamAssasin failures: 'spamd' bottleneck after update Thank you.0
Please sign in to leave a comment.
Comments
2 comments