-
Notifications
You must be signed in to change notification settings - Fork 832
Open
Labels
Description
Bug Summary
Cleaning up removed entries failed during normal operation
Description
I noticed the following log entries today while updating to 2.2.14 (the entries are from before the update):
Oct 25 12:13:34 host miniflux[1534]: level=ERROR msg="Unable to cleanup removed entries" user_id=1 feed_id=209 error="store: unable to cleanup entries: pq: deadlock detected"
Oct 25 12:13:34 host miniflux[1534]: level=INFO msg="Clearing content from removed entries completed" removed_entries_content_cleared=2420
The PostgreSQL log during the same timeframe contains the following:
2025-10-25 12:13:34.019 CEST [163325] ERROR: deadlock detected
2025-10-25 12:13:34.019 CEST [163325] DETAIL: Process 163325 waits for ShareLock on transaction 468296164; blocked by process 163310.
Process 163310 waits for ShareLock on transaction 468296234; blocked by process 163325.
Process 163325:
DELETE FROM
entries
WHERE
feed_id=$1 AND
status=$2 AND
NOT (hash=ANY($3))
Process 163310:
UPDATE
entries
SET
title='',
content=NULL,
url='',
author=NULL,
comments_url=NULL,
document_vectors=NULL
WHERE id IN (
SELECT id
FROM entries
WHERE status = $1 AND content IS NOT NULL
ORDER BY id ASC
LIMIT $2
)
2025-10-25 12:13:34.019 CEST [163325] HINT: See server log for query details.
2025-10-25 12:13:34.019 CEST [163325] CONTEXT: while deleting tuple (2467,6) in relation "entries"
2025-10-25 12:13:34.019 CEST [163325] STATEMENT:
DELETE FROM
entries
WHERE
feed_id=$1 AND
status=$2 AND
NOT (hash=ANY($3))
Steps to Reproduce
No idea, sorry
Expected Behavior
No deadlock should occur
Actual Behavior
A deadlock occurs
Version
2.2.13
Browser
No response
Relevant Logs or Error Output
Additional Context
No response
Checklist
- I have searched existing issues to ensure this bug hasn't been reported before.