[Mageia-sysadm] [Mageia-dev] servers downtime

[Michael Scherer]

Le mardi 11 janvier 2011 à 15:20 +0100, nicolas vigier a écrit :
> On Tue, 11 Jan 2011, Michael Scherer wrote:
> 
> > 
> > Just to be complete, what will be impacted are the server in marseille
> > datacenter :
> > - valstar ( svn, main scheduler, ldap )
> > - ecosse, buildhost
> > - jonund, buildhost
> > - alamut ( dns, new mailling list server, web application server for
> > everything except blog and website )
> 
> All those servers are now rebooted with updated kernel, thanks to TMB.

Alamut got broken.

I have seen the issue on my vm too :
- sshd start
- httpd start
- sympa start, stop because postgresql is not here
- postgresql is not run ( as there is no dependency on it )
- puppet ( who would start pgsql ) is not run

-> bad

so I connect by ssh, I start postgresql, I restart sympa => boot
continue.

Obvious solution, adding postgresql in the init file :

# Required-Start: $network mysql

Better solution, patch the script or sympa so sympa do not block it
Maybe the switch to systemd will do it too.

Anyone care to send a bug report to mdv ? 

-- 
Michael Scherer