You are not logged in.
Pages: 1
I have been trying to start a deployment of OCS for our Mac clients here, but although it has been installed, the 2 test machines have not checked in to the server. The one that does work seems to check in at some arbitrary interval that I cannot find the settings for.
I have already altered the org.ocs.agent.plist file to run every 120 secs, and while it does run, it does not check in every time, it only checks in when it wants. I'd like to know how to get it to do a check in at a interval of my choosing. (our windows system check in every 2 hours). I've scoured every file I can think of to look for when it chooses to check in.
thanks much
The normal result of the 120 sec cycle is :
[Fri Aug 14 11:25:46 2009][debug] Ocsinventory unified agent for UNIX and Linux0.0.9.2cvs
[Fri Aug 14 11:25:46 2009][debug] Log system initialised
[Fri Aug 14 11:25:46 2009][debug] the --server passed doesn't have a protocle, assume http as default
[Fri Aug 14 11:25:46 2009][debug] Accountinfo file: /var/lib/ocsinventory-agent/http:__{servername}_ocsinventory/ocsinv.adm
[Fri Aug 14 11:25:47 2009][debug] A TAG seems to already exist in the server for thismachine. If so, the -t paramter is useless. Please change the TAGdirectly on the server.
[Fri Aug 14 11:25:47 2009][debug] Time to call Proc::Daemon
[Fri Aug 14 11:25:47 2009][debug] Daemon started
[Fri Aug 14 11:25:47 2009][debug] Proc::PID::File avalaible, checking for pid file
[Fri Aug 14 11:25:47 2009][debug] parent process already exists
this differs from the times when it does check in :
[Thu Aug 13 09:26:04 2009][debug] Ocsinventory unified agent for UNIX and Linux0.0.9.2cvs
[Thu Aug 13 09:26:04 2009][debug] Log system initialised
[Thu Aug 13 09:26:04 2009][debug] the --server passed doesn't have a protocle, assume http as default
[Thu Aug 13 09:26:04 2009][debug] Accountinfo file: /var/lib/ocsinventory-agent/http:__{servername}_ocsinventory/ocsinv.adm
[Thu Aug 13 09:26:04 2009][debug] A TAG seems to already exist in the server for thismachine. If so, the -t paramter is useless. Please change the TAGdirectly on the server.
[Thu Aug 13 09:26:04 2009][debug] Time to call Proc::Daemon
[Thu Aug 13 09:26:04 2009][debug] Daemon started
[Thu Aug 13 09:26:04 2009][debug] Proc::PID::File avalaible, checking for pid file
[Thu Aug 13 09:26:04 2009][debug] parent process already exists
[Thu Aug 13 09:27:24 2009][debug] Calling handlers : `start_handler'
[Thu Aug 13 09:27:24 2009][debug] Compress::Zlib is avalaible.
[Thu Aug 13 09:27:24 2009][debug] sending XML
[Thu Aug 13 09:27:24 2009][debug] Calling handlers : `prolog_writers'
[Thu Aug 13 09:27:24 2009][debug] =BEGIN=SERVER RET======
[Thu Aug 13 09:27:24 2009][debug] $VAR1 = '<?xml version=\'1.0\' encoding=\'ISO-8859-1\'?>
{Misc windows stuff from the server clipped out here}
[Thu Aug 13 09:27:24 2009][debug] =END=SERVER RET======
[Thu Aug 13 09:27:24 2009][info] PROLOG_FREQ has not changed since last process
[Thu Aug 13 09:27:24 2009][debug] Calling handlers : `prolog_reader'
[Thu Aug 13 09:27:24 2009][debug] run func: `Ocsinventory::Agent::Option::Download::download_prolog_reader'
[Thu Aug 13 09:27:24 2009][info] Don't send the inventory
[Thu Aug 13 09:27:24 2009][debug] Calling handlers : `end_handler'
[Thu Aug 13 09:27:24 2009][debug] run func: `Ocsinventory::Agent::Option::Download::download_end_handler'
[Thu Aug 13 09:27:24 2009][info] Going to sleep for 2818 second(s)
Last edited by mrlint (2009-08-14 16:37:32)
Offline
We are not using the unified agent but are using CodeJanitor's version so this may no apply but from our experience with CodeJanitor's version the MAC agent only calls the server to upload inventory or changes to inventory at computer start up (or at reboot) which in our case is usually first thing in the morning unless someone has to reboot for some reason. We also have many faculty that do not turn their MACs off so inventory dates may not show today as last inventory date (because they have not turned their computer off or reboot for days).
Offline
Dear Mr. Lint,
I think your problem is that 120 seconds is *way* too short an interval. The message "Proc::PID::File avalaible, checking for pid file" and "parent process already exists" seem to imply that you are restarting the daemon before the last version has quit. On my system it takes 2:40 (160 seconds) to run. If you have it still running it will conflict. You should be able to run this with the debug version to check if that is what is going on.
Please post your launchd plist since I had to heavily modify mine to get a reliable run pattern. I set my plist file to update upon boot or every 24 hours. It seems to be reliable.
Offline
Hi,
I have exactly the same problem on my mac client, with the same log with "parent process already exist",and nothing is coming on my server (nothing in the logs on the server)
If you find a solution, please let me know,
regards
Last edited by dolmac (2010-07-12 10:08:41)
Offline
Pages: 1