Load Warnings on Slave

[bigdigillc]

I have asked a similar question before, but now there appears to be a new twist. I first was seeing about 6 processes running for logrotate each taking up about 1gig in memory. It appears it was being caused by maldet which I removed. Now I am getting bombarded with load warning and have over 50 instances of sisinfoc.php running and over 50 instances of httpd running. Nothing is hosted on this server and it only acts as a MySQL server. I have attached images. Please help. Thanks!!

[MRatWork]

Look like using ruid2 or itk or suphp instead other with _event/_worker postfix for 'php type'. Use 'php-fpm_event' and then select 'medium' for 'apache optimize'.

[bigdigillc]

I have the server set to "none" for Web server. Nothing is hosted tbere. Only Mysql.

[MRatWork]

Run 'chkconfig httpd off; service httpd stop' in slave.

[bigdigillc]

Thanks. That made the httpd processes go away, but now logrotate is back hogging resources and I get emails every minute or so saying 'Action: sh /script/restart-web --force >/dev/null 2>&1'

[MRatWork]

Uncheck 'status' for web in 'watchdog' for slave server.

[bigdigillc]

Ok, Thank you and I apologize, but I have the DNS restart notice now. I want to keep DNS on this slave, but it's not letting me switch to another DNS. It just sits for a long time then gives me an internal server 500 error. Is there a way to do it from the command line?

[MRatWork]

What's dns server in slave?.

[bigdigillc]

It's currently stuck on BIND but it was PDNS.

[bigdigillc]

Any thoughts?

[MRatWork]

Try in slave:

Code: [Select]

yum remove bind bind-utils -y
pkill -9 named
yum install pdns pdns-backend-mysql pdns-tools pdns-geo -y
sh /script/restart-dns -y

[bigdigillc]

Thanks, I got this.

Code: [Select]

[root@slave1 ~]# yum remove bind bind-utils -y
Loaded plugins: fastestmirror, protectbase, replace
Setting up Remove Process
No Match for argument: bind
Loading mirror speeds from cached hostfile
Could not retrieve mirrorlist http://download.webmin.com/download/yum/mirrorlist error was
[Errno 4] IOError: <urlopen error (110, 'Connection timed out')>
Error: Cannot find a valid baseurl for repo: Webmin


[MRatWork]

Try next an hour. It's because something wrong with webmin repo (maybe update in progress).

[bigdigillc]

Still getting this on slave server. What do I do? It's eating all my systems memory.

Code: [Select]

12000 ? 1195.52 root /usr/sbin/logrotate /etc/logrotate.conf
7684 ? 1096.57 root /usr/sbin/logrotate /etc/logrotate.conf
5290 ? 1047.29 root /usr/sbin/logrotate /etc/logrotate.conf
11569 ? 998.42 root /usr/sbin/logrotate /etc/logrotate.conf
4371 ? 951.45 root /usr/sbin/logrotate /etc/logrotate.conf
17687 ? 907.16 root /usr/sbin/logrotate /etc/logrotate.conf
9141 ? 884.48 root /usr/sbin/logrotate /etc/logrotate.conf
1584 ? 857.71 root /usr/sbin/logrotate /etc/logrotate.conf
20823 ? 829.37 root /usr/sbin/logrotate /etc/logrotate.conf
30620 ? 779.94 root /usr/sbin/logrotate /etc/logrotate.conf


[MRatWork]

Look like logrotate running via cron. Look like something trouble in logrotate. Try reboot your slave.

To disable logrotate, remove/delete '/etc/cron.daily/logrotate'.