note: JOBid limit is mentioned here: https://github.com/SchedMD/slurm/blob/master/RELEASE_NOTES <https://github.com/SchedMD/slurm/blob/master/RELEASE_NOTES>
greets Ron > On 6 Mar 2018, at 15:53, Ron Golberg <ron.golb...@icloud.com> wrote: > > Hi all, > I would like to implement Slurm in my current HPC system. > I have many Jobs divided into job arrays - which makes me cross the Slurm’s > 67 Million JOBuid limit. > I've looked into the source code and it looks like the ID’s are being reused > (67 Mil jobs cycle) but Slurm can handle identical IDs with the help of > another UID in the accounting DB (called: db_index). > So I understand that I can submit more than 67 Million Jobs, but is it > possible to use the real unique ID from the accounting db for Slurm > operations? > for example - check job status. > will it work if I won’t use accounting db at all? > > also, I have an external application that manages jobs sent to the scheduler, > Is it ok for it to rely on the db_index for managing jobs on Slurm (is the > accounting db always up-to date)? > > Thanks in advance, > Ron > > >
