Bug #54554: Bluestore volume selector improperly tracks bluefs log size - bluestore - Ceph

Actions

Copy link

Bug #54554

closed

Bluestore volume selector improperly tracks bluefs log size

Added by Igor Fedotov about 2 years ago. Updated over 1 year ago.

Status:

Resolved

Priority:

Normal

Assignee:

Target version:

% Done:

Source:

Tags:

Backport:

Regression:

Severity:

3 - minor

Reviewed:

Affected Versions:

Ceph - v16.2.7

ceph-qa-suite:

Pull request ID:

45408

Crash signature (v1):

Crash signature (v2):

Description

Here is the output of "bluefs stats" admin command:

ceph tell osd.1 bluefs stats

1 : device size 0x3fffe000 : using 0x23400000(564 MiB)
2 : device size 0x1900000000 : using 0x8c000(560 KiB)
RocksDBBlueFSVolumeSelector: wal_total:0, db_total:1020054732, slow_total:102005473280, db_avail:0
Usage matrix:
DEV/LEV WAL DB SLOW * * REAL FILES
LOG 0 B 110 MiB 0 B 0 B 0 B 2.1 MiB 1
WAL 0 B 550 MiB 0 B 0 B 0 B 549 MiB 1
DB 0 B 9 MiB 0 B 0 B 0 B 76 KiB 10
SLOW 0 B 0 B 0 B 0 B 0 B 0 B 0
TOTALS 0 B 669 MiB 0 B 0 B 0 B 0 B 12

And compare to bluefs perf counters dump:
"bluefs": {
"log_bytes": 4542464,
"log_compactions": 6,
"logged_bytes": 95711232,

Looks like bluefs stats tracking lacks counter decrement and shows amount of logged (i.e. processed) bytes instead as a result.