Hi,
I've changed the queues on lcgce02 to gridAMDCE02 and gridIntelCE02
(added 'CE02'), which fixed the problems for lcgce02 but not lcgce01.
Only after doing a similar change for the queues on ce01, all Nagios
tests turned green.
I am hoping that all ambiguity is now resolved. For the sake of the
WMS registered queues, I am changing the lcgce01 queues back to the
original state. Let's see how the system reacts to it.
It definitely was a learning experience, thank you for your help!
Cheers,
Luke
On 2 June 2014 18:25, Andrew Lahiff <[log in to unmask]> wrote:
>
> We're currently using a cluster node with both CREAM and ARC CEs (*), but I think once the CREAM CEs are decommissioned we will try to find a way to decommission the cluster node as well.
>
> Regards,
> Andrew.
>
> (*) the cluster node is publishing clusters corresponding to each generation of worker node, while each ARC CE has its own cluster associated with it.
>
> ________________________________________
> From: Stephen Burke [[log in to unmask]]
> Sent: Monday, June 02, 2014 5:52 PM
> To: [log in to unmask]
> Subject: Re: Debugging 'BrokerHelper: no compatible resources'.
>
> L Kreczko [mailto:[log in to unmask]] said:
>> Thanks for the comments. Stephen, what do you mean by 'cluster node
>> type'? I could not find a similar parameter in the ARC CE
>> documentation.
>
> Sorry, I missed the fact that this was with ARC CEs. In principle the cluster node could probably be adapted fairly easily to work with ARC, but not out of the box.
>
> Stephen
> --
> Scanned by iCritical.
> --
> Scanned by iCritical.
--
*********************************************************
Dr Lukasz Kreczko +44 (0)117 928 8724
CMS Group
School of Physics
University of Bristol
*********************************************************
|