Ceph

10/29/2019

07:44 AM RADOS Bug #40583: Lower the default value of osd_deep_scrub_large_omap_object_key_threshold

I am taking the liberty to add a couple of recent mailing list threads here that highlight a potentially unintended c... Florian Haas

10/23/2019

02:33 PM Dashboard Bug #42445 (Resolved): Dashboard can't handle self-signed cert on Grafana API

It appears that the dashboard is very unhappy if it is configured with a Grafana API URL that uses HTTPS and a self-s... Florian Haas

10/22/2019

02:32 PM RADOS Bug #41191: osd: scrub error on big objects; make bluestore refuse to start on big objects

Thanks! It looks like the backport for Mimic has already landed while the one for Nautilus is still pending. That str... Florian Haas

10/10/2019

08:36 AM RADOS Bug #41255: backfill_toofull seen on cluster where the most full OSD is at 1%

Adding 14.2.4 as an affected version, as I am seeing the same issue on a 14.2.4 cluster that has recently had 9 OSDs ... Florian Haas

10/08/2019

08:02 AM RADOS Bug #41191: osd: scrub error on big objects; make bluestore refuse to start on big objects

As far as I can see, without this patch there is no way to detect excessively large objects other than doing @rados s... Florian Haas

08/27/2019

01:19 PM RADOS Bug #40029: ceph-mon: Caught signal (Aborted) in (CrushWrapper::update_choose_args(CephContext*)+0x2fa) [0x7f516505614a]

Should add one more thing: the only clusters bitten by this issue would be those that, *at any time,* ran the @balanc... Florian Haas

08/26/2019

11:23 AM RADOS Bug #40029: ceph-mon: Caught signal (Aborted) in (CrushWrapper::update_choose_args(CephContext*)+0x2fa) [0x7f516505614a]

With thanks to Paul Emmerich in https://lists.ceph.io/hyperkitty/list/ceph-users@ceph.io/thread/QGY75UVQEAT2SUHHKZC2K... Florian Haas

08/22/2019

09:53 AM RADOS Bug #40029: ceph-mon: Caught signal (Aborted) in (CrushWrapper::update_choose_args(CephContext*)+0x2fa) [0x7f516505614a]

We just bumped into this, on Luminous (12.2.12). It actually caused us momentary loss of quorum.
Sequence of event... Florian Haas

08/09/2019

12:57 PM RADOS Backport #39694: luminous: _txc_add_transaction error (39) Directory not empty not handled on operation 21 (op 1, counting from 0)

Just adding a note here to say that we've encountered this problem (or a very similar one, see https://tracker.ceph.c... Florian Haas

12:52 PM RADOS Bug #38724: _txc_add_transaction error (39) Directory not empty not handled on operation 21 (op 1, counting from 0)

So we're seeing a similar error after upgrading a Jewel cluster to Luminous (12.2.12), and then adding BlueStore OSDs... Florian Haas