# Killing jobs using qdel -Wforce does not fire EXECJOB\_PRETERM/EXECJOB\_EPILOGUE events **URL:** \ **Category:** Developers\ **Created:** [May 10, 2022, 3:05pm UTC](https://community.openpbs.org/t/killing-jobs-using-qdel-wforce-does-not-fire-execjob-preterm-execjob-epilogue-events/3086 "2022-05-10T15:05:09Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![ziggycc](https://avatars.discourse-cdn.com/v4/letter/z/ed655f/32.png) [@ziggycc](https://community.openpbs.org/u/ziggycc)\ **Post date:** [May 10, 2022, 3:05pm UTC](https://community.openpbs.org/t/killing-jobs-using-qdel-wforce-does-not-fire-execjob-preterm-execjob-epilogue-events/3086/1 "2022-05-10T15:05:09Z")
Hi all, I wrote a hook script to manage the local scratch space on the compute nodes. Particularly, in case of multiple nodes exclusive jobs, the scratch space is created on the fly. Such scratch space is virtually composed by the all local scratches on the involved nodes and it is managed by the master node only. If such multiple nodes exclusive jobs are killed by using qdel -Wforce , pbs.EXECJOB\_PRETERM and pbs.EXECJOB\_EPILOGUE events do not run and consequently I am not able to properly free the “distrubuted” scratch space. Using pbs.EXECHOST\_PERIODIC event could be quite complicated. Do you know any events that fires in this particular situation or any workarounds? Thanks in advance for helping. ---
**Author:** ![Source](https://yyz2.discourse-cdn.com/flex030/user_avatar/community.openpbs.org/source/32/756_2.png) [@Source](https://community.openpbs.org/u/Source)\ **Post date:** [July 17, 2023, 8:29am UTC](https://community.openpbs.org/t/killing-jobs-using-qdel-wforce-does-not-fire-execjob-preterm-execjob-epilogue-events/3086/2 "2023-07-17T08:29:34Z")
I think its intention, qdel -W force will not trigger any clean up, and it would remove the database item regardless the really ‘kill’ occurs or not . Correct me if i misunderstood it