Hi Adrian,
Yes, you if-else synopsis is correct; though it's a little had to capture
that linger is temporary (i.e. it only applies the specified number of
times).
Andy
On Fri, 14 Jul 2017, Adrian Sevcenco wrote:
> On 07/13/2017 12:27 AM, Andrew Hanushevsky wrote:
>> Hi Adrian,
> Hi!
>
>> Let's say you want to create a file and server x was selected. Then you
>> want to create another file and the cms looks and says well, all things are
>> roughly equal who should I select? If linger is set to 1 then server x will
>> be selected. If linger is zero or server x was already selected twice in a
>> row, then another server would be selected. The cms.sche has no impact on
>> this behaviour.
> i see! so,
> if cms.sched priorities are equal for all servers {
> if (linger)
> { will keep the previous use server for writing }
> else
> { choose another server }
> else
> { chose highest priority }
>
> did i understood correctly?
>
>> We put this in way back and no one uses it as far as I know because it's
>> really hard to predict the space allocation pattern and people generally
>> don't like adding extra entropy to the decision making.
> :)) yes, i can understand that :))
>
> Thank you!
> Adrian
>
>
>
>>
>> Andy
>>
>> On Wed, 12 Jul 2017, Adrian Sevcenco wrote:
>>
>>> Hi! It is not clear to me what does cms.space linger do ...
>>> the documentation says :
>>> "The number of times a server may be reselected without an intervening
>>> server being selected for allocation. The default is zero (0)."
>>>
>>> but what is "an intervening server" ?
>>>
>>> also, how this interact with the scheduling done by cms.sched?
>>>
>>> Thank you!
>>> Adrian
>>>
>>>
>>> ########################################################################
>>> 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
>>>
>>
>
>
> --
> ----------------------------------------------
> Adrian Sevcenco, Ph.D. |
> Institute of Space Science - ISS, Romania |
> adrian.sevcenco at {cern.ch,spacescience.ro} |
> ----------------------------------------------
>
>
########################################################################
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
|