On 24/10/2022 09:32, Ole Holm Nielsen wrote:
On 10/24/22 06:12, Richard Chang wrote:I have a two node Slurmctld setup and both will mount an NFS exported directory as the state save location.It is definitely a BAD idea to store Slurm StateSaveLocation on a slow NFS directory! SchedMD recommends to use local NVME or SSD disks because there will be many IOPS to this file system! I recommend you to read "Field Notes 6: From The Frontlines of Slurm Support", Jason Booth, SchedMD available from https://slurm.schedmd.com/publications.html. Read the Hardware pages 18-20 which recommend: Fast path to the StateSaveLocation ■ IOPS this filesystem can sustain is a major bottleneck to job throughput ● At least 2 directories and two files created per job ● The corresponding unlink() calls will add to the load
So what is the recommended way if you want to have HA with slurmctld? Ward
smime.p7s
Description: S/MIME Cryptographic Signature
