2024.1.1-p2 Release Notes

Resolved Issues

Internal Number Products Case Number Description
VOV-16906 Accelerator CS0549375 Fixed an issue whereby a reservation record was being deleted from the system in the situation where Monitor had dropped a License-based resource at the time of the reservation.
VOV-16976 Accelerator CS0562250 Corrected an issue where the osname field of the tasker vovselect object was not identified correctly for some RedHat based Linux distros, including RHEL 9.
VOV-17052 Allocator None Resolved a bug where the .csv export button in the Allocator web UI would output a blank file. Now, the .csv export button exports the data properly.
VOV-17208 All CS0591663 To address an issue with patching an Arm installation, it is important that the 2023.1.2-p5.scripts.patch.tar.gz is applied before the other 2023.1.2-p5 patch files.
VOV-17255 Accelerator AAP20698, AAP21947, AAP23994 Addressed conflict where running multiple MPI jobs in the background could create filename conflicts resulting in job failure.
VOV-17264 Accelerator Plus CS0598688 Restored ability to specify Job placement policy for Accelerator Plus with <SWD>/jobclass/<queuename>.tcl.
VOV-17297 Accelerator Plus CS0601565 Fixed bug where Accelerator Plus could show incorrect values when units other than MB are used on resource values such as RAM/1GB.
VOV-17314 Accelerator CS0615890 Enabled container support for the top component job of a DP job submission. The top component job will now honor the container specified in the primary resource request in the submission (for example, -r Container:foo).
VOV-17377 Accelerator Plus CS0615719 Fixed a bug that prevented Accelerator Plus or FlowTracer vovwxd, version 2024.1.0 and higher, from submitting job arrays to base queues running older versions (2023.1.2 and lower).
VOV-17383 Accelerator, Accelerator Plus CS0615679 Fix for bsub that now properly keys off either VOV_JOB_DESC(snapProp) or VOV_JOB_DESC(env) correctly before attempting SNAPPROP setup, and doesn't execute unnecessary code if either is not set.
VOV-17390 Accelerator Plus None Fixed a race condition that could prevent taskers from fast exiting on a heavily loaded server. STARTTIME and MAXIDLE are implemented as fields of the tasker object so are no longer accessible as properties. An alternate, faster connect protocol for taskers that do not already have a corresponding tasker object in the server may be enabled by using the -K command line option or by setting VOV_WX_DD_QUICKCONNECT=1 in the Accelerator Plus Direct Drive tasker environment.
VOV-17395 Accelerator Plus CS0600818 Fixed an issue whereby the Accelerator Plus web UI Ext Link was pointing to an invalid base queue URL. The invalid, or incomplete, URL was based on the short host name rather than a fully qualified host name.
VOV-17403 Accelerator None Handled tasker exit issue on non-parsable signature for container hook. Fixed interactive job shell stuck issue.
VOV-17405 Accelerator None Fixed resource miscounting issue on wrong signature for enter hook container.
VOV-17406 Accelerator None Fixed issue in Ctrl-C handling in the setup container hook flow letting Accelerator jobs to continue to run.
VOV-17414 Accelerator Plus None Fixed a rare race condition where a job can end and its tasker be shut down and the server processes the shut down message prior to the job end message. In this scenario, the tasker log would show that the job exited cleanly but the job status would be failed due to the tasker being killed. The job status will now reflect the correct status. Fixed feature that allows vovserver to defer the processing of lower-priority client requests when under stress.
VOV-17460 Accelerator Plus None vovwxd now postpones the log cleaner execution at startup, hence improving the startup performance A new, CONFIG(cleanerPauseInterval) parameter has been added to control the wait time after processing each file, which helps to reduce disk I/O load, improving overall system performance.
VOV-17561 All CS0643188 Fixed issue where NUMA node configuration that does not contain any CPUs could crash vovtasker.
VOV-17594 None None Fix incorrect initialization of wxagent autokill values occasionally seen on vovwxd restart.
VOV-17597 Monitor CS0647276 Resolved a bug where the .csv export button in the Monitor web UI would output a blank file. Now, the .csv export button exports the data properly.
VOV-18081 Accelerator, Accelerator Plus None Fixed a crash in the tasker when it receives a job as it is exiting.
VOV-18082 None None Fixed an issue that caused stalled buckets in certain scenarios with buckets refilling after deletion in the base queue and tasker failures.
VOV-18083 Accelerator Plus None Fixed a problem where vovwxd with Direct Drive setup could crash due to a race condition when multiple base queues were in use.