Newer
Older
This file describes changes in recent versions of Slurm. It primarily
documents those changes that are of interest to users and administrators.
* Changes in Slurm 14.11.9
==========================
-- Correct "sdiag" backfill cycle time calculation if it yields locks. A
microsecond value was being treated as a second value resulting in an
overflow in the calcuation.
-- Fix segfault when updating timelimit on jobarray task.
-- Fix to job array update logic that can result in a task ID of 4294967294.
-- Fix of job array update, previous logic could fail to update some tasks
of a job array for some fields.
-- CRAY - Fix seg fault if a blade is replaced and slurmctld is restarted.
* Changes in Slurm 14.11.8
==========================
-- Eliminate need for user to set user_id on job_update calls.
-- Correct list of unavailable nodes reported in a job's "reason" field when
that job can not start.
-- Map job --mem-per-cpu=0 to --mem=0.
-- Fix squeue -o %m and %d unit conversion to Megabytes.
-- Fix issue with incorrect time calculation in the priority plugin when
a job runs past it's time limit.
-- Prevent users from setting job's partition to an invalid partition.
-- Fix sreport core dump when requesting
'job SizesByAccount grouping=individual'.
-- select/linear: Correct count of CPUs allocated to job on system with
hyperthreads.
Brian Christiansen
committed
-- Fix race condition where last array task might not get updated in the db.
-- CRAY - Remove libpmi from rpm install
-- Fix squeue -o %X output to correctly handle NO_VAL and suffix.
-- When deleting a job from the system set the job_id to 0 to avoid memory
corruption if thread uses the pointer basing validity off the id.
-- Fix issue where sbatch would set ntasks-per-node to 0 making any srun
afterward cause a divide by zero error.
-- switch/cray: Refine logic to set PMI_CRAY_NO_SMP_ENV environment variable.
-- When sacctmgr loads archives with version less than 14.11 set the array
task id to NO_VAL, so sacct can display the job ids correctly.
-- When using memory cgroup if a task uses more memory than requested
the failures are logged into memory.failcnt count file by cgroup
and the user is notified by slurmstepd about it.
-- Fix scheduling inconsistency with GRES bound to specific CPUs.
-- If user belongs to a group which has split entries in /etc/group
search for its username in all groups.
-- Do not consider nodes explicitly powered up as DOWN with reason of "Node
unexpected rebooted".
-- Use correct slurmd spooldir when creating cpu-frequency locks.
-- Note that TICKET_BASED fairshare will be deprecated in the future. Consider
using the FAIR_TREE algorithm instead.
-- Set job's reason to BadConstaints when job can't run on any node.
-- Prevent abort on update of reservation with no nodes (licenses only).
-- Prevent slurmctld from dumping core ifjob_resrcs is missing in the
job data structure.
-- Fix squeue to print array task ids according to man page when
SLURM_BITSTR_LEN is defined in the environment.
-- In squeue sort jobs based on array job ID if available.
-- Fix the calculation of job energy by not including the NO_VAL values.
-- Advanced reservation fixes: enable update of bluegene reservation, avoid
abort on multi-core reservations.
-- Set the totalview_stepid to the value of the job step instead of NO_VAL.
-- Fix slurmdbd core dump if the daemon does not have connection with
the database.
-- Display error message when attempting to modify priority of a held job.
-- Backfill scheduler: The configured backfill_interval value (default 30
seconds) is now interpretted as a maximum run time for the backfill
scheduler. Once reached, the scheduler will build a new job queue and
start over, even if not all jobs have been tested.
-- Backfill scheduler now considers OverTimeLimit and KillWait configuration
parameters to estimate when running jobs will exit.
-- Correct task layout with CR_Pack_Node option and more than 1 CPU per task.
-- Fix the scontrol man page describing the release argument.
-- When job QOS is modified, do so before attempting to change partition in
order to validate the partition's Allow/DenyQOS parameter.
* Changes in Slurm 14.11.7
==========================
-- Initialize some variables used with the srun --no-alloc option that may
cause random failures.
-- Add SchedulerParameters option of sched_min_interval that controls the
minimum time interval between any job scheduling action. The default value
is zero (disabled).
-- Change default SchedulerParameters=max_sched_time from 4 seconds to 2.
-- Refactor scancel so that all pending jobs are cancelled before starting
cancellation of running jobs. Otherwise they happen in parallel and the
pending jobs can be scheduled on resources as the running jobs are being
cancelled.
-- ALPS - Add new cray.conf variable NoAPIDSignalOnKill. When set to yes this
will make it so the slurmctld will not signal the apid's in a batch job.
Instead it relies on the rpc coming from the slurmctld to kill the job to
end things correctly.
-- ALPS - Have the slurmstepd running a batch job wait for an ALPS release
before ending the job.
-- Initialize variables in consumable resource plugin to prevent core dump.
-- Fix scancel bug which could return an error on attempt to signal a job step.
-- In slurmctld communication agent, make the thread timeout be the configured
value of MessageTimeout rather than 30 seconds.
-- sshare -U/--Users only flag was used uninitialized.
-- Cray systems, add "plugstack.conf.template" sample SPANK configuration file.
-- BLUEGENE - Set DB2NOEXITLIST when starting the slurmctld daemon to avoid
random crashing in db2 when the slurmctld is exiting.
-- Make full node reservations display correctly the core count instead of
cpu count.
-- Preserve original errno on execve() failure in task plugin.
-- Add SLURM_JOB_NAME env variable to an salloc's environment.
-- Overwrite SLURM_JOB_NAME in an srun when it gets an allocation.
-- Make sure each job has a wckey if that is something that is tracked.
-- Make sure old step data is cleared when job is requeued.
-- Load libtinfo as needed when building ncurses tools.
-- Fix small memory leak in backup controller.
-- Fix segfault when backup controller takes control for second time.
-- Cray - Fix backup controller running native Slurm.
-- Provide prototypes for init_setproctitle()/fini_setproctitle on NetBSD.
-- Add configuration test to find out the full path to su command.
-- preempt/job_prio plugin: Fix for possible infinite loop when identifying
preemptable jobs.
-- preempt/job_prio plugin: Implement the concept of Warm-up Time here. Use
the QoS GraceTime as the amount of time to wait before preempting.
Basically, skip preemption if your time is not up.
-- Make srun wait KillWait time when a task is cancelled.
-- switch/cray: Revert logic added to 14.11.6 that set "PMI_CRAY_NO_SMP_ENV=1"
if CR_PACK_NODES is configured.
* Changes in Slurm 14.11.6
==========================
-- If SchedulerParameters value of bf_min_age_reserve is configured, then
a newly submitted job can start immediately even if there is a higher
priority non-runnable job which has been waiting for less time than
bf_min_age_reserve.
-- qsub wrapper modified to export "all" with -V option
-- RequeueExit and RequeueExitHold configuration parameters modified to accept
numeric ranges. For example "RequeueExit=1,2,3,4" and "RequeueExit=1-4" are
equivalent.
-- Correct the job array specification parser to accept brackets in job array
expression (e.g. "123_[4,7-9]").
-- Fix for misleading job submit failure errors sent to users. Previous error
could indicate why specific nodes could not be used (e.g. too small memory)
when other nodes could be used, but were not for another reason.
-- Fix squeue --array to display correctly the array elements when the
% separator is specified at the array submission time.
-- Fix priority from not being calculated correctly due to memory issues.
-- Fix a transient pending reason 'JobId=job_id has invalid QOS'.
-- A non-administrator change to job priority will not be persistent except
for holding the job. User's wanting to change a job priority on a persistent
basis should reset it's "nice" value.
-- Print buffer sizes as unsigned values when failed to pack messages.
Brian Christiansen
committed
-- Fix race condition where sprio would print factors without weights applied.
-- Document the sacct option JobIDRaw which for arrays prints the jobid instead
of the arrayTaskId.
-- Allow users to modify MinCPUsNode, MinMemoryNode and MinTmpDiskNode of
their own jobs.
-- Increase the jobid print field in SQUEUE_FORMAT in
opt_modulefiles_slurm.in.
-- Enable compiling without optimizations and with debugging symbols by
default. Disable this by configuring with --disable-debug.
-- job_submit/lua plugin: Add mail_type and mail_user fields.
-- Use standard statvfs(2) syscall if available, in preference to
non-standard statfs.
-- Add a new option -U/--Users to sshare to display only users
information, parent and ancestors are not printed.
-- Purge 50000 records at a time so that locks can released periodically.
-- Fix potentially uninitialized variables
-- ALPS - Fix issue where a frontend node could become unresponsive and never
added back into the system.
-- Gate epilog complete messages as done with other messages
-- If we have more than a certain number of agents (50) wait longer when gating
rpcs.
-- FrontEnd - ping non-responding or down nodes.
-- switch/cray: If CR_PACK_NODES is configured, then set the environment
variable "PMI_CRAY_NO_SMP_ENV=1"
-- Fix invalid memory reference in SlurmDBD when putting a node up.
-- Allow opening of plugstack.conf even when a symlink.
-- Fix scontrol reboot so that rebooted nodes will not be set down with reason
'Node xyz unexpectedly rebooted' but will be correctly put back to service.
-- CRAY - Throttle the post NHC operations as to not hog the job write lock
if many steps/jobs finish at once.
-- Disable changes to GRES count while jobs are running on the node.
-- CRAY - Fix issue with scontrol reconfig.
-- slurmd: Remove wrong reporting of "Error reading step ... memory limit".
The logic was treating success as an error.
-- Eliminate "Node ping apparently hung" error messages.
-- Fix average CPU frequency calculation.
-- When allocating resources with resolution of sockets, charge the job for all
CPUs on allocated sockets rather than just the CPUs on used cores.
-- Prevent slurmdbd error if cluster added or removed while rollup in progress.
Removing a cluster can cause slurmdbd to abort. Adding a cluster can cause
the slurmdbd rollup to hang.
-- sview - When right clicking on a tab make sure we don't display the page
list, but only the column list.
-- FRONTEND - If doing a clean start make sure the nodes are brought up in the
database.
-- MySQL - Fix issue when using the TrackSlurmctldDown and nodes are down at
the same time, don't double bill the down time.
-- MySQL - Various memory leak fixes.
-- Fix node manager logic to keep unexpectedly rebooted node in state
NODE_STATE_DOWN even if already down when rebooted.
-- Fix for array jobs submitted to multiple partitions not starting.
-- CRAY - Enable ALPs mpp compatibility code in sbatch for native Slurm.
-- ALPS - Move basil_inventory to less confusing function.
Loading
Loading full blame...