false alerts on process resources

[Attila Heidrich]

Hi!

RHQ 4.9

Sometimes we have got "status down" alerts of process resources. The
processes haven't got any problems meanwhile, and we don't have the false
positives for each and every process on the same platform.

The processes are mainly "vboxheadless something"-like processes - at lease
we only have problems with these. The alerts are cleared very soon, but
sthey are till really annoying.

Anyone having the same problem, or preferably a solution?

Regards,

Attila

[Thomas Segismont]

Le 30/01/2014 12:28, Attila Heidrich a écrit :
> Hi!
>
> RHQ 4.9
>
> Sometimes we have got "status down" alerts of process resources. The
> processes haven't got any problems meanwhile, and we don't have the
> false positives for each and every process on the same platform.
>
> The processes are mainly "vboxheadless something"-like processes - at
> lease we only have problems with these. The alerts are cleared very
> soon, but sthey are till really annoying.
>
> Anyone having the same problem, or preferably a solution?
>
> Regards,
>
> Attila
>
> rhq-users mailing list
> [e-mail]
> https://lists.fedorahosted.org/mailman/listinfo/rhq-users

Hi Attila,

Just to understand if it's a problem with alerting or with process 
resources availability reports: does the process resource show some down 
availability at the time of the alert?

Can you set the org.rhq.plugins.platform.ProcessComponent logger to 
debug level? You should then be able to see exceptions in the agent log, 
if they occur during avail collection.

Thanks,
Thomas

[Attila Heidrich]

Hi!

Thanks the quick answer!

The resource really shows the downtime, the alert and the "reverse alert"
also received - but there's nothing in the VirtualBox logs what confirm the
downtime. Also the guest OS send the logs with syslog-ng while its VBox
process is reported to be down.

How can I set the debug level (for the given component)?

Attila

2014-01-30 Thomas Segismont
> Le 30/01/2014 12:28, Attila Heidrich a écrit :
> > Hi!
> >
> > RHQ 4.9
> >
> > Sometimes we have got "status down" alerts of process resources. The
> > processes haven't got any problems meanwhile, and we don't have the
> > false positives for each and every process on the same platform.
> >
> > The processes are mainly "vboxheadless something"-like processes - at
> > lease we only have problems with these. The alerts are cleared very
> > soon, but sthey are till really annoying.
> >
> > Anyone having the same problem, or preferably a solution?
> >
> > Regards,
> >
> > Attila
> >
> > rhq-users mailing list
> > [e-mail]
> > https://lists.fedorahosted.org/mailman/listinfo/rhq-users
>
> Hi Attila,
>
> Just to understand if it's a problem with alerting or with process
> resources availability reports: does the process resource show some down
> availability at the time of the alert?
>
> Can you set the org.rhq.plugins.platform.ProcessComponent logger to debug
> level? You should then be able to see exceptions in the agent log, if they
> occur during avail collection.
>
> Thanks,
> Thomas
> _______________________________________________
> rhq-users mailing list
> [e-mail]
> https://lists.fedorahosted.org/mailman/listinfo/rhq-users

[Thomas Segismont]

FAQ !!! :p

https://docs.jboss.org/author/display/RHQ/FAQ#FAQ-HowdoIgetdebugmessagesfrom...

Le 30/01/2014 13:36, Attila Heidrich a écrit :
> Hi!
>
> Thanks the quick answer!
>
> The resource really shows the downtime, the alert and the "reverse
> alert" also received - but there's nothing in the VirtualBox logs what
> confirm the downtime. Also the guest OS send the logs with syslog-ng
> while its VBox process is reported to be down.
>
> How can I set the debug level (for the given component)?
>
> Attila
>
> 2014-01-30 Thomas Segismont <[e-mail]
> >
> Le 30/01/2014 12:28, Attila Heidrich a écrit :
>
>     Hi!
>
>     RHQ 4.9
>
>     Sometimes we have got "status down" alerts of process resources. The
>     processes haven't got any problems meanwhile, and we don't have the
>     false positives for each and every process on the same platform.
>
>     The processes are mainly "vboxheadless something"-like processes
>     - at
>     lease we only have problems with these. The alerts are cleared very
>     soon, but sthey are till really annoying.
>
>     Anyone having the same problem, or preferably a solution?
>
>     Regards,
>
>     Attila
>
>
>     _________________________________________________
>     rhq-users mailing list
>     [e-mail].__org
>     <mailto:[e-mail]>
>     https://lists.fedorahosted.__org/mailman/listinfo/rhq-users
>     <https://lists.fedorahosted.org/mailman/listinfo/rhq-users>
>
>
> Hi Attila,
>
> Just to understand if it's a problem with alerting or with process
> resources availability reports: does the process resource show some
> down availability at the time of the alert?
>
> Can you set the org.rhq.plugins.platform.__ProcessComponent logger
> to debug level? You should then be able to see exceptions in the
> agent log, if they occur during avail collection.
>
> Thanks,
> Thomas
> _________________________________________________
> rhq-users mailing list
> [e-mail].__org
> <mailto:[e-mail]>
> https://lists.fedorahosted.__org/mailman/listinfo/rhq-users
> <https://lists.fedorahosted.org/mailman/listinfo/rhq-users>
>
>
>
>
> rhq-users mailing list
> [e-mail]
> https://lists.fedorahosted.org/mailman/listinfo/rhq-users

[Jay Shaughnessy]

This does not seem like an issue with alerting.  The resource is being 
reported with DOWN availability and so you get the alert. (Same with UP, 
I guess).  So it really is a question of why DOWN was reported if, as 
you say, the resource seems to be UP the entire time.  It must have to 
do with the component's inquiry into the Process component...

On 1/30/2014 7:36 AM, Attila Heidrich wrote:
> Hi!
>
> Thanks the quick answer!
>
> The resource really shows the downtime, the alert and the "reverse 
> alert" also received - but there's nothing in the VirtualBox logs what 
> confirm the downtime. Also the guest OS send the logs with syslog-ng 
> while its VBox process is reported to be down.
>
> How can I set the debug level (for the given component)?
>
> Attila
>
> 2014-01-30 Thomas Segismont <[e-mail] 
> >
> Le 30/01/2014 12:28, Attila Heidrich a écrit :
>
>     Hi!
>
>     RHQ 4.9
>
>     Sometimes we have got "status down" alerts of process
>     resources. The
>     processes haven't got any problems meanwhile, and we don't
>     have the
>     false positives for each and every process on the same platform.
>
>     The processes are mainly "vboxheadless something"-like
>     processes - at
>     lease we only have problems with these. The alerts are cleared
>     very
>     soon, but sthey are till really annoying.
>
>     Anyone having the same problem, or preferably a solution?
>
>     Regards,
>
>     Attila
>
>
>     _______________________________________________
>     rhq-users mailing list
>     [e-mail]
>     <mailto:[e-mail]>
>     https://lists.fedorahosted.org/mailman/listinfo/rhq-users
>
>
> Hi Attila,
>
> Just to understand if it's a problem with alerting or with process
> resources availability reports: does the process resource show
> some down availability at the time of the alert?
>
> Can you set the org.rhq.plugins.platform.ProcessComponent logger
> to debug level? You should then be able to see exceptions in the
> agent log, if they occur during avail collection.
>
> Thanks,
> Thomas
> _______________________________________________
> rhq-users mailing list
> [e-mail]
> <mailto:[e-mail]>
> https://lists.fedorahosted.org/mailman/listinfo/rhq-users
>
>
>
>
> rhq-users mailing list
> [e-mail]
> https://lists.fedorahosted.org/mailman/listinfo/rhq-users