[slurm-users] Is It Possible to change the node order for different partition

Ryan Novosielski novosirj at rutgers.edu
Wed Jun 27 08:32:50 MDT 2018


Bear in mind that at least, last I knew, using a topology plugin will cause all of this to be ignored and, at least in our case, for SLURM to do roughly the opposite of what you want regarding weights (for us, this is because the highest weighted equipment was connected to the switch with the least connected equipment, so it tended to aim any job that would fit there to those nodes).

> On Jun 27, 2018, at 12:42 AM, Brian Andrus <toomuchit at gmail.com> wrote:
> 
> Yes, if you put a weight parameter to the nodes.
> 
> From the manpage:
> 
> Weight
> 
>     The priority of the node for scheduling purposes. All things being equal, jobs will be allocated the nodes with the lowest weight which satisfies their requirements. For example, a heterogeneous collection of nodes might be placed into a single partition for greater system utilization, responsiveness and capability. It would be preferable to allocate smaller memory nodes rather than larger memory nodes if either will satisfy a job's requirements. The units of weight are arbitrary, but larger weights should be assigned to nodes with more processors, memory, disk space, higher processor speed, etc. Note that if a job allocation request can not be satisfied using the nodes with the lowest weight, the set of nodes with the next lowest weight is added to the set of nodes under consideration for use (repeat as needed for higher weight values). If you absolutely want to minimize the number of higher weight nodes allocated to a job (at a cost of higher scheduling overhead), give each node a distinct Weight value and they will be added to the pool of nodes being considered for scheduling individually. The default value is 1.
> 
> Brian Andrus
> 
> 
> On 6/26/2018 3:06 PM, Bill wrote:
>> Hi Everyone,
>> 
>> For example, I have two partitions, high,low each has same nodes node[1-10], When we submit job to high partition the nodes order is node1,node2......node10,  when we submit job to low partition, the order is node10,node9......node1.
>> 
>> Is it possible to do that?
>> 
>> Thanks in advance,
>> Bill
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: Message signed with OpenPGP
URL: <http://lists.schedmd.com/pipermail/slurm-users/attachments/20180627/32d0e470/attachment-0001.sig>


More information about the slurm-users mailing list