Manual Check - State and count of processes (don't match process)

[Davide]

Hi All!
Since 1.6 I was able to match nginx process which PS is:

apache   14296 24146  0 03:47 ?        00:00:20 nginx: worker process                                                              
apache   14297 24146  0 03:47 ?        00:05:01 nginx: worker process                                                              
apache   14298 24146  0 03:47 ?        00:01:11 nginx: worker process                                                              
apache   14299 24146  0 03:47 ?        00:00:28 nginx: worker process                                                              
apache   14300 24146  0 03:47 ?        00:00:40 nginx: worker process                                                              
apache   14301 24146  0 03:47 ?        00:01:50 nginx: worker process                                                              
apache   14303 24146  0 03:47 ?        00:01:11 nginx: worker process                                                              
apache   14304 24146  0 03:47 ?        00:02:52 nginx: worker process                                                              
apache   14306 24146  0 03:47 ?        00:00:00 nginx: cache manager process                                                       
root     24146     1  0 Apr23 ?        00:00:00 nginx: master process /usr/local/openresty/nginx/sbin/nginx -c /usr/local/openresty/nginx/conf/nginx.conf

by this simple regex:
(/usr)?(/local)?(/bin/|/sbin/)?(/)?nginx

By updating to 2.0 the check doesn’t work anymore.

I tryed these regex:

• (/usr)?(/local)?(/bin/|/sbin/)?(/)?nginx(.*)$
• nginx

also I tryed with “exact name of the process without arguments”

• nginx
• nginx:

No one of above is worked.

The agent is still 1.6 but AFAIK it shouldn’t matter, right?

[Davide]

Trying to update agent the Check goes from CRIT to UNKNW with output:
Item not found in monitoring data

cmk -d web-mint01 | grep -e Version -e ps -e nginx
AGENT OUTPUT:

Version: 2.0.0p3
<<<ps_lnx>>>
root 0 0 00:00:00 110-08:59:27 168 [kpsmoused]
root 189172 3392 00:00:00 110-08:56:47 2155 cupsd -C /etc/cups/cupsd.conf
apache 224096 12580 00:00:21 19:50:09 14296 nginx: worker process 
apache 226152 15000 00:05:08 19:50:09 14297 nginx: worker process 
apache 224888 13688 00:01:14 19:50:09 14298 nginx: worker process 
apache 224896 13932 00:00:30 19:50:09 14299 nginx: worker process 
apache 226508 14924 00:00:42 19:50:09 14300 nginx: worker process 
apache 226636 16032 00:01:52 19:50:09 14301 nginx: worker process 
apache 225276 14120 00:01:12 19:50:09 14303 nginx: worker process 
apache 231188 18172 00:02:56 19:50:09 14304 nginx: worker process 
apache 203712 2816 00:00:00 19:50:09 14306 nginx: cache manager process 
root 201204 6288 00:00:00 18-14:33:14 24146 nginx: master process /usr/local/openresty/nginx/sbin/nginx -c /usr/local/openresty/nginx/conf/nginx.conf
root 122680 1144 00:00:00 00:01 30161 ps ax -o user:32,vsz,rss,cputime,etime,pid,command --columns 10000
pswpin 0
pswpout 0
thp_collapse_alloc 68288
thp_collapse_alloc_failed 0

[slummm (Jesus Henao)]

Hi

Aparently it’s a bug in 2.0.0p3 version of check_mk Server.

Update from that to 2.0.0p4 works fine for me. At check_mk server.

In the monitored host, you can use rather p3 or p4 version of the agent

[Davide]

Works also with ealier version of agent (1.6.0p17)!

[slummm (Jesus Henao)]

Hi,

I’m not sure, we came from 1.6.0p, and in 2.0.0p3 doesn’t works fine.

We upgrade both, agent (at the client) and Server and same time.

But you can try in a POC client monitoring any process like httpd, mysqld, if you have an error, it would be present with any process.

I hope this was helpfully to you

[Davide]

Hi,
after upgrade the server to 2.0.0p4 I managed to monitoring any process either for earlier versions of the agent. So the “bug” is present in the 2.0.0p3 server side.