Here's a twist.

All my current WU's (~ 150 WU's) just disappeared from BOINCMgr after aborting a WU and this is the error log:

2019-05-02 02:43:06 AM | Milkyway@Home | task de_modfit_85_bundle4_4s_south4s_0_1556550902_938938_0 aborted by user
2019-05-02 02:43:45 AM | Milkyway@Home | update requested by user
2019-05-02 02:43:46 AM | Milkyway@Home | Sending scheduler request: Requested by user.
2019-05-02 02:43:46 AM | Milkyway@Home | Reporting 1 completed tasks
2019-05-02 02:43:46 AM | Milkyway@Home | Requesting new tasks for AMD/ATI GPU
2019-05-02 02:43:48 AM | Milkyway@Home | Scheduler request completed: got 0 new tasks
2019-05-02 02:43:48 AM | Milkyway@Home | Not sending work - last request too recent: 83 sec
2019-05-02 02:43:48 AM | Milkyway@Home | [error] garbage_collect(); still have active task for acked result de_modfit_85_bundle4_4s_south4s_0_1556550902_938938_0; state 5

This is the third time in the last few days I've seen a group of WU's disappear.

I've been testing the OC capability of the Sapphire 280x and having some driver crashes, but never seen an entire GPU WU queue just disappear after a driver crash.

The remaining related entries are:

2019-05-02 02:43:55 AM | Milkyway@Home | Computation for task de_modfit_85_bundle4_4s_south4s_0_1556550902_938938_0 finished
2019-05-02 02:45:21 AM | Milkyway@Home | Sending scheduler request: To fetch work.
2019-05-02 02:45:21 AM | Milkyway@Home | Requesting new tasks for AMD/ATI GPU
2019-05-02 02:45:24 AM | Milkyway@Home | Scheduler request completed: got 300 new tasks