Mon Oct 18, 2021 11:44 pm
Login Register Lost Password? Contact Us


Multiple ECLAgent processes

Questions related to node architecture, redundancy and system monitoring

Tue May 14, 2013 3:18 pm Change Time Zone

Hi,

I noticed that somehow my system has spun up many eclagent processes. Is this normal? My version is 3.10.4-1. A list of all running processes under the hpcc user:

Code: Select all
  PID TTY          TIME CMD
2384 ?        00:00:10 eclagent
2673 ?        00:00:10 eclagent
2697 ?        00:00:10 eclagent
2834 ?        00:00:10 eclagent
2846 ?        00:00:10 eclagent
2982 ?        00:00:10 eclagent
3033 ?        00:00:11 eclagent
3183 ?        00:00:10 eclagent
3326 ?        00:00:10 eclagent
3476 ?        00:00:10 eclagent
3607 ?        00:00:10 eclagent
3760 ?        00:00:10 eclagent
3931 ?        00:00:10 eclagent
4149 ?        00:00:10 eclagent
4302 ?        00:00:10 eclagent
4451 ?        00:00:10 eclagent
4614 ?        00:00:10 eclagent
4777 ?        00:00:10 eclagent
4924 ?        00:00:10 eclagent
5074 ?        00:00:10 eclagent
5230 ?        00:00:10 eclagent
5369 ?        00:00:10 eclagent
5528 ?        00:00:10 eclagent
5667 ?        00:00:10 eclagent
5838 ?        00:00:10 eclagent
5980 ?        00:00:10 eclagent
5992 ?        00:00:10 eclagent
6125 ?        00:00:10 eclagent
6137 ?        00:00:10 eclagent
6274 ?        00:00:11 eclagent
6286 ?        00:00:10 eclagent
6812 ?        00:00:10 eclagent
7335 ?        00:00:10 eclagent
7857 ?        00:00:10 eclagent
10732 ?        00:00:00 init_dali
10764 ?        07:52:38 daserver
10872 ?        00:00:00 init_dfuserver
10905 ?        00:04:50 dfuserver
11017 ?        00:00:00 init_eclagent
11050 ?        00:00:08 agentexec
11151 ?        00:00:00 init_eclccserve
11183 ?        00:00:10 eclccserver
11284 ?        00:00:00 init_eclschedul
11316 ?        00:00:06 eclscheduler
11420 ?        00:00:00 init_esp
11452 ?        00:13:38 esp
11555 ?        00:00:00 init_roxie
11588 ?        13:52:41 roxie
11696 ?        00:00:00 init_sasha
11731 ?        00:00:42 saserver
11937 ?        00:00:00 init_thor
11971 ?        00:00:00 start_thor
12067 ?        00:00:00 run_thor
14767 ?        00:00:00 init_dafilesrv
14796 ?        00:26:20 dafilesrv
29073 ?        00:02:30 thormaster_myth
sbagaria
Community Advisory Board Member
Community Advisory Board Member
 
Posts: 92
Joined: Thu Jul 05, 2012 9:23 pm

Wed May 15, 2013 2:10 pm Change Time Zone

Hello Sid, this could be normal. The way it works is that the agentexec process spawns an ECLAGENT instance for each active workunit (eithor THOR or HTHOR). When the workunits complete, the eclagent process terminates. So, if you open ECLWatch and look at the "Activity" or "Browse Workunits" screen, it is likely that you will see there are active workunits.
william.whitehead@lexisnexis.com
 
Posts: 3
Joined: Fri Apr 29, 2011 12:41 pm

Wed May 15, 2013 2:16 pm Change Time Zone

Thanks for clarifying that. I don't have any active workunits running, so this may very well be zombie processes which failed to abort. BUT their CPU time has increased since yesterday (from 10 seconds to 11 seconds) which means they are lightweight processes still active and not showing up in the activity page. I don't have anything scheduled either, so they can not be polling threads used by the scheduler. Is there any logfile I can examine to check for active processes? I would like to know what these processes are doing.
sbagaria
Community Advisory Board Member
Community Advisory Board Member
 
Posts: 92
Joined: Thu Jul 05, 2012 9:23 pm

Wed May 15, 2013 2:42 pm Change Time Zone

Yes, you can look at the eclagent logfiles, which are normally located in /var/log/HPCCSystems/myeclagent. Feel free to attach it to this thread and I can have a look as well
william.whitehead@lexisnexis.com
 
Posts: 3
Joined: Fri Apr 29, 2011 12:41 pm


Return to System Health

Who is online

Users browsing this forum: No registered users and 1 guest

cron