8 replies to this topic

[Joseph Orendain]

Hello All.

Is there a way to determine what causes the PRO4BUSD daemon to be killed on our unix server? Is there a log to check in the unix side and maybe in the concerto side? Please advise.

Thanks a lot!

Joseph

[Rick Young]

I've never seen pro4busd "disappear" by itself - normally (if that's not oxymoronic in this example), if it "dies" it just hangs out there, and needs to be killed. Thus, I'd probably be blaming a human...

If you (as root) look at /.sh_history (or possibly equivalent file dependent on shell(s) used or 'nix flavour), you may be able to see someone issuing a kill command.

Once again depending on 'nix flavour, paths may vary: /var/adm/sulog would show a normal user "su'ing" to root, then I believe the history commands are logged to /home/username/.sh_history

HTH
Rick

[Rob Donovan]

Hi,

Our Bus has also died for apparently no reason in the past, but only a couple of times, so we havent looked into it much.

If I notice anything, I'll keep you informed...

Rob.

[Chris Mackenzie]

Hello All.

Is there a way to determine what causes the PRO4BUSD daemon to be killed on our unix server? Is there a log to check in the unix side and maybe in the concerto side? Please advise.

Thanks a lot!

Joseph

Hi,

if you have P4EVENTREPORTLEVEL=4 in the pro4 bus section
of pro4.ini then messages will be written to syslog (/var/adm/logs/syslog or suchlike)

[Joseph Orendain]

thank you to all those who replied.

I checked our syslog when our daemon died and here's the message :

Dec 6 11:26:16 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:33:14 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:34:14 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:35:17 erpprod monitor.sh: CHECK_POINT
Dec 6 11:35:12 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:35:54 erpprod bootpd[2572]: Option 43: Client specified unacceptable option length (200 octets)
Dec 6 11:40:58 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:41:10 erpprod syslog: Unrecognised error number

Dec 6 11:50:31 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:51:03 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:51:48 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10


What does the error mean? Do we have to change a setting in our kernel? Please advise.

Thanks!

Joseph

Hello All.

Is there a way to determine what causes the PRO4BUSD daemon to be killed on our unix server? Is there a log to check in the unix side and maybe in the concerto side? Please advise.

Thanks a lot!

Joseph

Hi,

if you have P4EVENTREPORTLEVEL=4 in the pro4 bus section
of pro4.ini then messages will be written to syslog (/var/adm/logs/syslog or suchlike)

[Rick Young]

In your /etc/pro4.ini and/or in your alternate .ini file note in the [Enivornments] section (I think) of pro4.ini, that is being called in your Bus & Tasks api call, there'll be some parameters. (Off the top of my head without a manual or .ini file handy) MAXKERNELS=##, COMPONENT_USERS=##. These are documented in the Bus & Tasks manual if I recall correctly.

That error message would indicate that you have either 10 maximum kernels, period, or you have 10 kernels reserved for Bus & Tasks usage - and you are having > 10 concurrent Bus calls being made. I've not experienced this killing the bus daemon (which is not to say that it cannot happen), hence, I'd perhaps be wondering if any of the Tasks are "going rogue" or not terminating, or not terminating correctly and this perhaps is trashing the daemon.

You may find additional info by taking Chris' suggestion of adding the P4EVENTREPORTLEVEL and some of the other parameters for tracing (once again, don't have these off the top of my head, but I know they exist elsewhere in the forum).



-Rick

[Chris Mackenzie]

I checked our syslog when our daemon died and here's the message :
Dec 6 11:35:12 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
Dec 6 11:35:54 erpprod bootpd[2572]: Option 43: Client specified unacceptable option length (200 octets)
Dec 6 11:40:58 erpprod syslog: Pro-IV Bus Daemon - the Maximum Kernel Pool Limit is set to 10
What does the error mean? Do we have to change a setting in our kernel? Please advise.

Thanks!

Joseph


Joseph

Looks like one of your calls into the API is passing parameters that do not
match with the defined task.


You may want to increase the number of available connections as well
if 10 is not enought. Debugging this stuff is _such_ fun!

[Joseph Orendain]

Thanks All!

When I change values of parameters in /etc/pro4_64.ini, do I have to re-start the daemon? I am not sure if newly created session will then pick-up the updated values. Please advise.

Thanks!

Joseph

[Rick Young]

...sometimes...

But, it is probably advisable to