mirror of https://github.com/ghostfolio/ghostfolio
Browse Source
A portfolio snapshot computation that runs longer than the Bull queue's lock duration is flagged as stalled and fails, surfacing as HTTP 500 on all portfolio and account endpoints. Because the snapshot job uses the user id as its job id and was only configured with removeOnComplete, the failed job stayed in the queue and every later request re-read it, making the 500 permanent until Redis was flushed by hand. - Add removeOnFail to PORTFOLIO_SNAPSHOT_PROCESS_JOB_OPTIONS so a failed job is removed and the next request recomputes instead of returning a cached failure - Increase the default portfolio snapshot computation timeout from 30s to 300s so large portfolios compute successfully out of the box (still overridable via PROCESSOR_PORTFOLIO_SNAPSHOT_COMPUTATION_TIMEOUT) Refs #6914, #3755pull/6916/head
2 changed files with 5 additions and 2 deletions
Loading…
Reference in new issue