[Mageia-sysadm] [sysadmin-reports] Hobbit [817408] alamut.mageia.org:hobbitd warning (YELLOW)

Michael Scherer misc at zarb.org
Mon Jan 16 17:25:41 CET 2012


Le lundi 16 janvier 2012 à 16:59 +0100, nicolas vigier a écrit :
> On Mon, 16 Jan 2012, Michael Scherer wrote:
> 
> > Le lundi 16 janvier 2012 à 14:10 +0200, Buchan Milne a écrit :
> > > Why are we restarting xymon every hour?
> > > 
> > 
> > because the initscripts do not have a 'status' command and so xymon is
> > restarted by puppet ( in fact, not restarted, just started ). 
> > I supposed that xymon initscript would be idempotent and just do
> > nothing, and since you were aware of the issue last time we talk, I
> > supposed it was not a problem beyong "1 useless line of log". 
> > 
> > But if this is the reason why xymon is flooding the list ( and i didn't
> > figured this was the problem after looking at the code ), I will take a
> > look to avoid that. Puppet can surely work around this.
> 
> Yes, hasstatus parameter to xymon service can be set to false if
> init script does not support "status" :
> http://docs.puppetlabs.com/references/stable/type.html#service

Yep, but after it fail since it try to loopup in the process table for a
process called xymon ( who doesn't exist ). 

So we need to add our own status command, which thankfully was already
done by the script used to start xymon. 

I pushed this on svn, see commit r2281 and r2279. 

-- 
Michael Scherer



More information about the Mageia-sysadm mailing list