Limit pg log length during recovery/backfill so that we don't run out of memory.
This means if there's another failure, we'll need to restart backfill or go from recovery to backfill, but that's better than running out of memory.
Treating osd_max_pg_log_entries as a hard cap should be sufficient.
- Source set to Development
- Backport set to luminous
- Assignee changed from David Zafman to Josh Durgin
- Assignee changed from Josh Durgin to Neha Ojha
- Backport changed from luminous to mimic, luminous
- Status changed from 12 to Fix Under Review
- Status changed from Fix Under Review to Pending Backport
- Copied to Backport #24988: luminous: Limit pg log length during recovery/backfill so that we don't run out of memory. added
- Copied to Backport #24989: mimic: Limit pg log length during recovery/backfill so that we don't run out of memory. added
- Related to Bug #25198: FAILED assert(trim_to <= info.last_complete) in PGLog::trim() added
- Related to Bug #26868: PGLog.cc: saw valgrind issues while accessing complete_to->version added
- Status changed from Pending Backport to Resolved
Also available in: Atom