<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class="">
We basically always do this. Just be mindful of how long it takes to upgrade your database (if you have that ability to do a dry run, you might ant to do that). That’s true of any upgrade, though.
<div class=""><br class="">
</div>
<div class="">If you have to skip more than one version, you’ll have to upgrade in stages.<br class="">
<div class=""><br class="">
<div>
<blockquote type="cite" class="">
<div class="">On Nov 10, 2022, at 7:00 PM, Michael Gutteridge <<a href="mailto:michael.gutteridge@gmail.com" class="">michael.gutteridge@gmail.com</a>> wrote:</div>
<br class="Apple-interchange-newline">
<div class="">
<div dir="ltr" class="">
<div class="gmail_default" style="font-family:monospace">Theoretically I think you should be able to. Slurm should upgrade from the previous two releases (see
<a href="https://slurm.schedmd.com/quickstart_admin.html#upgrade:~:text=Slurm%20permits%20upgrades%20to%20a%20new%20major%20release%20from%20the%20past%20two%20major%20releases%2C" class="">
this</a>) and I think that should include 20.11. (20.11 -> 21.08 -> 22.05). Not something I've done though.</div>
<div class="gmail_default" style="font-family:monospace"><br class="">
</div>
<div class="gmail_default" style="font-family:monospace"> - Michael</div>
<div class="gmail_default" style="font-family:monospace"><br class="">
</div>
</div>
<br class="">
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Thu, Nov 10, 2022 at 2:15 PM Sid Young <<a href="mailto:sid.young@gmail.com" class="">sid.young@gmail.com</a>> wrote:<br class="">
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div dir="ltr" class="">
<div dir="ltr" class="">Is there a direct upgrade path from <span style="font-size:12pt;font-family:"Times New Roman",serif" class="">
20.11.0 to 22.05.6 or is it in multiple steps?</span><br clear="all" class="">
<div class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div dir="ltr" class="">
<div class=""><br class="">
</div>
<div class=""><br class="">
</div>
<div class="">Sid Young</div>
<div class=""><br class="">
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<br class="">
</div>
<br class="">
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Fri, Nov 11, 2022 at 7:53 AM Marshall Garey <<a href="mailto:marshall@schedmd.com" target="_blank" class="">marshall@schedmd.com</a>> wrote:<br class="">
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
We are pleased to announce the availability of Slurm version 22.05.6.<br class="">
<br class="">
This includes a fix to core selection for steps which could result in <br class="">
random task launch failures, alongside a number of other moderate <br class="">
severity issues.<br class="">
<br class="">
- Marshall<br class="">
<br class="">
--<br class="">
Marshall Garey<br class="">
Release Management, Support, and Development<br class="">
SchedMD LLC - Commercial Slurm Development and Support<br class="">
<br class="">
> * Changes in Slurm 22.05.6<br class="">
> ==========================<br class="">
> -- Fix a partition's DisableRootJobs=no from preventing root jobs from working.<br class="">
> -- Fix the number of allocated cpus for an auto-adjustment case in which the<br class="">
> job requests --ntasks-per-node and --mem (per-node) but the limit is<br class="">
> MaxMemPerCPU.<br class="">
> -- Fix POWER_DOWN_FORCE request leaving node in completing state.<br class="">
> -- Do not count magnetic reservation queue records towards backfill limits.<br class="">
> -- Clarify error message when --send-libs=yes or BcastParameters=send_libs<br class="">
> fails to identify shared library files, and avoid creating an empty<br class="">
> "<filename>_libs" directory on the target filesystem.<br class="">
> -- Fix missing CoreSpec on dynamic nodes upon slurmctld restart.<br class="">
> -- Fix node state reporting when using specialized cores.<br class="">
> -- Fix number of CPUs allocated if --cpus-per-gpu used.<br class="">
> -- Add flag ignore_prefer_validation to not validate --prefer on a job.<br class="">
> -- Fix salloc/sbatch SLURM_TASKS_PER_NODE output environment variable when the<br class="">
> number of tasks is not requested.<br class="">
> -- Permit using wildcard magic cookies with X11 forwarding.<br class="">
> -- cgroup/v2 - Add check for swap when running OOM check after task<br class="">
> termination.<br class="">
> -- Fix deadlock caused by race condition when disabling power save with a<br class="">
> reconfigure.<br class="">
> -- Fix memory leak in the dbd when container is sent to the database.<br class="">
> -- openapi/dbv0.0.38 - correct dbv0.0.38_tres_info.<br class="">
> -- Fix node SuspendTime, SuspendTimeout, ResumeTimeout being updated after<br class="">
> altering partition node lists with scontrol.<br class="">
> -- jobcomp/elasticsearch - fix data_t memory leak after serialization.<br class="">
> -- Fix issue where '*' wasn't accepted in gpu/cpu bind.<br class="">
> -- Fix SLURM_GPUS_ON_NODE for shared GPU gres (MPS, shards).<br class="">
> -- Add SLURM_SHARDS_ON_NODE environment variable for shards.<br class="">
> -- Fix srun error with overcommit.<br class="">
> -- Fix bug in core selection for the default cyclic distribution of tasks<br class="">
> across sockets, that resulted in random task launch failures.<br class="">
> -- Fix core selection for steps requesting multiple tasks per core when<br class="">
> allocation contains more cores than required for step.<br class="">
> -- gpu/nvml - Fix MIG minor number generation when GPU minor number<br class="">
> (/dev/nvidia[minor_number]) and index (as seen in nvidia-smi) do not match.<br class="">
> -- Fix accrue time underflow errors after slurmctld reconfig or restart.<br class="">
> -- Surpress errant errors from prolog_complete about being unable to locate<br class="">
> "node:(null)".<br class="">
> -- Fix issue where shards were selected from multiple gpus and failed to<br class="">
> allocate.<br class="">
> -- Fix step cpu count calculation when using --ntasks-per-gpu=.<br class="">
> -- Fix overflow problems when validating array index parameters in slurmctld<br class="">
> and prevent a potential condition causing slurmctld to crash.<br class="">
> -- Remove dependency on json-c in slurmctld when running with power saving.<br class="">
> Only the new "SLURM_RESUME_FILE" support relies on this, and it will be<br class="">
> disabled if json-c support is unavailable instead.<br class="">
<br class="">
</blockquote>
</div>
</div>
</blockquote>
</div>
</div>
</blockquote>
</div>
<br class="">
</div>
</div>
</body>
</html>