Describe the bug
A PH5 contained Array_t_001, 004, 005, and, due to some problems reading SEG-D headers, also Array_t_2228668, 2589373, 4784829, 6292028. (The problems reading SEG-D headers were fixed before the PH5 was completed.) The user building it, and myself, both attempted delete_table -n master.ph5 --all_arrays but even though a backup table was written the 4 Array_t with abnormally high values did not delete from the master
Environment (please complete the following information):
User machine (likely Linux) and PH5 server at EPIC, delete_table 2021.336
To Reproduce
Run delete_table -n master.ph5 --all_arrays on the specified master.ph5 on the EPIC server. The messages '829 Not found', '28 Not found', '668 Not found', and '373 Not found' should appear in stdout but not in nuke_table.log
Expected behavior
All Array_t to delete with --all_arrays flag. However, this case is extremely unusual because it requires 1) a problem reading the SEG-D headers, 2) that problem to not cause segdtoph5 to fail, and 3) time considerations favoring continuing with the PH5 instead of rebuilding completely.
Additional context
Using delete_table -n master.ph5 -A 2228668, etc. was an effective workaround.
hrotman-pic
commented
1 month ago
Describe the bug A PH5 contained Array_t_001, 004, 005, and, due to some problems reading SEG-D headers, also Array_t_2228668, 2589373, 4784829, 6292028. (The problems reading SEG-D headers were fixed before the PH5 was completed.) The user building it, and myself, both attempted delete_table -n master.ph5 --all_arrays but even though a backup table was written the 4 Array_t with abnormally high values did not delete from the master
Environment (please complete the following information): User machine (likely Linux) and PH5 server at EPIC, delete_table 2021.336
To Reproduce Run delete_table -n master.ph5 --all_arrays on the specified master.ph5 on the EPIC server. The messages '829 Not found', '28 Not found', '668 Not found', and '373 Not found' should appear in stdout but not in nuke_table.log
Expected behavior All Array_t to delete with --all_arrays flag. However, this case is extremely unusual because it requires 1) a problem reading the SEG-D headers, 2) that problem to not cause segdtoph5 to fail, and 3) time considerations favoring continuing with the PH5 instead of rebuilding completely.
Additional context Using delete_table -n master.ph5 -A 2228668, etc. was an effective workaround.