On 14/12/12 14:24, Daniela Bauer wrote:
> I think it's this one:
> https://wiki.italiangrid.it/twiki/bin/view/CREAM/KnownIssues#Issue_with_conflicting_BUpdaterS
>
" This will be fixed in EMI 2."
Assuming that Emyr isn't running EMI-1, then perhaps a GGUS ticket
against cream is warranted. Sometimes this does something. Sometimes it
just makes me feel better to have filed it...
> I think globus-gridftp might also need a (separate) restart.
Chris
>
> Cheers,
> Daniela
>
>
> On 14 December 2012 13:39, Matt Doidge <[log in to unmask]
> <mailto:[log in to unmask]>> wrote:
>
> Hello,
>
>
> There's a bug in /etc/init.d/gLite on the cream box whereby it
> gets the
> order wrong when it starts and stops so you see this....
>
> *** glite-ce-blahparser:
> Starting BNotifier: /usr/bin/BNotifier: Error creating and binding
> socket: Address already in use
> [FAILED]
>
>
> The bug wouldn't be in /etc/init.d/gLite, but in the config file it
> effectively sources, /etc/gLiteservices on EMI1 and
> /etc/gLiteservices.cream on EMI2. I have one of each type of cream
> running and notice the order is different in both. However I never
> "restart" my cream services, out of habit I stop, pray, then start them.
>
> However the last time I saw this type of error I had a hung
> BNotifier (sub)process, that required me to get killy with the PIDs
> to actually stop it. That was a while ago, on my glite cream IIRC,
> so things could be better now.
>
>
>
> Daniela (at least I think it was Daniela) sent me a link a while
> back on
> modifications to make to the file to sort this out. It involved
> piping a
> service list through sort.
>
>
> Daniela is a font of wisdom, to be honest I'm not sure what order
> things should be in the services file. My guess would be tomcat
> first, then the blah parser, then the lb-locallogger (with the order
> on these last two not being as important), but that doesn't fit with
> throwing a sort at the file.
>
> Now I'm confused, I though that these services needed to connect to
> cream (so cream, i.e. tomcat, should be started first), not the
> other way round.
>
>
>
> I can't seem to find the page anymore. Does this ring any bells with
> anyone ?
>
>
> My googling is failing me too.
>
> Sorry for not being as helpful as I'd like to be.
> Cheers,
> Matt
>
>
> Cheers,
>
> Emyr
>
>
>
>
> --
> Sent from the pit of despair
>
> -----------------------------------------------------------
> [log in to unmask] <mailto:[log in to unmask]>
> HEP Group/Physics Dep
> Imperial College
> Tel: +44-(0)20-75947810
> http://www.hep.ph.ic.ac.uk/~dbauer/
|