Hi Chris,
On 15 Jan 2013, at 00:28, Stephen Burke <[log in to unmask]> wrote:
> Christopher J. Walker [mailto:[log in to unmask]] said:
>>> Yes and no - it excludes CEs which don't have
>>> GlueCEStateStatus=Production, which would be the case if CEs in
>>> downtime are set to drain or are closed completely. However hepgrid10 is
>> still publishing Production ...
>>
>> So is that the bug?
>
> Which bug?
>
>> Wahid says[https://ggus.eu/ws/ticket_info.php?ticket=90391 ] of ce1 and
>> ce3 at ECDF:
>>
>> "ce1 and ce3 aren't being used anymore - they are not production in the
>> gocdb." And indeed they aren't, they are "not in production, not monitored
>> and not visible to EGI".
>
> Sites are responsible for what they publish. ce1 and ce3 are still in Edinburgh's site BDII and still have Status=Production, hence the WMS will try to submit there. If the CEs are supposed to be out of the production system they should be removed from the site BDII - the service-level flags in the GOC DB only affect the monitoring.
>
Yes this is more an inconsistency from our side. As far as I remember I held back on BDII removal whilst the CEs still existed in AGIS after decommissioning. There is no reason for these to be there now so I'll remove them from the site BDII - sorry for the confusion.
Cheers,
Andy.
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|