Closed hackartisan closed 1 month ago
Looking through some logs.
I've deleted the following jobs:
[{"enqueued_at"=>"2024-05-23T21:17:20Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/jennifer", "jid"=>"d16dd8bca8bfe846a877346b"},
{"enqueued_at"=>"2024-05-23T21:17:21Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/will", "jid"=>"3a5aebf70e005ac09c75131b"},
{"enqueued_at"=>"2024-05-23T21:17:20Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/jordan", "jid"=>"8859d74e543914a0e8eb420a"},
{"enqueued_at"=>"2024-05-23T21:17:21Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/eal_ncl", "jid"=>"77613df1c7e1ead4cfa5f6ee"},
{"enqueued_at"=>"2024-05-23T21:17:23Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/mary", "jid"=>"2b46722bbcd1549d318acb60"},
{"enqueued_at"=>"2024-05-23T21:17:20Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/squirrel", "jid"=>"0e9a35af7c16fb295fb9013d"},
{"enqueued_at"=>"2024-05-23T21:17:23Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/beth", "jid"=>"522bc0114e13311a98770ddf"},
{"enqueued_at"=>"2024-05-23T21:17:23Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/zuzanna", "jid"=>"98474e5be5c68f2739f9c0ee"},
{"enqueued_at"=>"2024-05-23T21:17:20Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/richel", "jid"=>"f39db2262d507c599b48ad67"},
{"enqueued_at"=>"2024-05-23T21:17:21Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/microfilm", "jid"=>"6506cd737235603dbce3cef2"},
{"enqueued_at"=>"2024-05-23T21:17:21Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/to_delete", "jid"=>"254da7d55b298dd5dd46a977"},
{"enqueued_at"=>"2024-05-23T21:17:23Z", "directory"=>"/mnt/hydra_sources/studio_new/DPUL/sherry", "jid"=>"9205a8abe506f74b7620c0f2"}]
I think the ingester deleted the resources that successfully made it through the ingest process. I think this is safe to close.
update: this morning right now there are 20,125 jobs in the default queue. super_low has > 88,000.
Summary
A bulk ingest was initiated for the studio_new directory. The resources created need to be deleted and the jobs need to be killed.
When I looked at the default queue this morning the jobs went up to
#26295
. the last couple are UpdateDao jobs. Before that I see jobs that look related to the large ingest, as well as a lot of derivative and characterization jobs, but we probably want to look more closely in the console.the high queue has 4 ocr request jobs enqueued. the low queue has 375 jobs enqueued, probably all preservation jobs. super_low has 42,758.
Workers were turned off manually via the UI around 3:52 am on Tuesday morning, 5/28. There may be ingests or other jobs enqueued after that time, which we should try not to delete.
Impact
Workers are currently off and until we can turn them back on figgy is essentially down for ingest.
Priority recommendation
Sudden Priority Justification
Nothing can go into figgy right now.