> On 11 Feb 2017, at 2:02 am, Allen Wittenauer <[email protected]> 
> wrote:
> 
> 
>> On Feb 9, 2017, at 6:29 PM, Gavin McDonald <[email protected]> wrote:
>> 
>> 1. The Hadoop Nodes H0-H9 - Hadoop and related projects have priority 
> 
>       I'll admit I'm not a Jenkins expert, so maybe I'm just missing it 
> but... how does the Jenkins job scheduler actually determine the job priority 
> for a given node?  I was under the impression our configuration was strictly 
> a label-based FIFO queue. In other words, if the queue looks like this 
> (lowest wait to highest wait ordering):
> 
> 
> hadoop 
> hbase
> wagon
> 
> all three have hadoop as a label, and the only node available was H0, that 
> wagon would get scheduled, not hbase or hadoop.
> 
>       If that isn't the case, how exactly does 'priority' work, especially 
> for e.g., H10 and all of the hadoop jobs now having Ubuntu as a label?

The labelling in the documentation I refer to is social. ‘Hadoop and related 
projects have priority’ means Hadoop uses these nodes first , and others later. 
‘Everybody is welcome’ means nobody has priority on these nodes.

Gav...

Reply via email to