Hi Marian,
I ushed a patch for this and it will be included in 4.2.4 which, I
believe, comes out in two weeks.
Andy
On Fri, 25 Sep 2015, Marian Zvada wrote:
> Thanks Andy, that nicely explains. We can live with that for now as we know
> it doesn't harm anything and understand this message as no real error. Using
> DNS alias is more and more popular in the federation setups, so if you can
> develop solution that preserve different cluster ID appearance it'll be well
> received by everyone I think.
>
> Thanks,
> Marian
>
> On 9/24/15 10:01 AM, Andrew Hanushevsky wrote:
>> Hi Marian,
>>
>> OK, I found the source of the problem. It's related to the fact that the
>> CERN DNS permutes the ist of IP addresses associated with
>> cms-xrd-global.cern.ch (it's load balanced). I thought I had taken care
>> of this problem but it appears that the code can be bypassed. This
>> causes a different cluster ID developed by each of the managers and the
>> meta-manager detects that and complains. I'll develop a new fix.
>> Generally, this creates more traffic on the network but otherwise should
>> not cause a problem in the way the redirector is being used. You could
>> get arround the problem by listing xrdcmsglobal01 and xrdcmsglobal02 as
>> separate entries (i.e. not use the global name). But it's probably not
>> necessary at this point.
>>
>> Andy
>>
>> On Tue, 22 Sep 2015, Marian Zvada wrote:
>>
>>> I forgot to say EU redirector restarted subscribes to the global ones
>>> mentioned in the log snipped...
>>>
>>> -Marian
>>>
>>> On 9/22/15 12:03 PM, Marian Zvada wrote:
>>>> Hi,
>>>>
>>>> Tom found this in his log files of redirector when restarting the
>>>> service:
>>>>
>>>> > 150922 09:59:44 31930 Manager: xrdcmsglobal02.cern.ch manager
>>>> configuration
>>>> > differs from xrdcmsglobal01.cern.ch for site local; making file
>>>> location
>>>> > unpredictable!
>>>>
>>>> What this might indicate? We've crosschecked config files on both global
>>>> redirectors and they are coherent!
>>>>
>>>> Thanks,
>>>> Marian
>>>>
>>>> ########################################################################
>>>> Use REPLY-ALL to reply to list
>>>>
>>>> To unsubscribe from the XROOTD-L list, click the following link:
>>>> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1
>>>
>>> ########################################################################
>>> Use REPLY-ALL to reply to list
>>>
>>> To unsubscribe from the XROOTD-L list, click the following link:
>>> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1
>>>
>
> ########################################################################
> Use REPLY-ALL to reply to list
>
> To unsubscribe from the XROOTD-L list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1
>
########################################################################
Use REPLY-ALL to reply to list
To unsubscribe from the XROOTD-L list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-L&A=1
|