Back to bug 1332083

Who When What Removed Added
Red Hat Bugzilla Rules Engine 2016-05-02 07:18:21 UTC Target Release 2.0 1.3.2
Vikhyat Umrao 2016-05-02 07:19:37 UTC Group redhat
Target Release 1.3.2 1.3.3
Samuel Just 2016-05-02 14:43:59 UTC Assignee sjust dzafman
Vikhyat Umrao 2016-05-03 04:27:07 UTC Status NEW ASSIGNED
Vikhyat Umrao 2016-05-06 07:19:34 UTC Flags needinfo?(dzafman)
David Zafman 2016-05-10 00:28:14 UTC CC bengland
Flags needinfo?(dzafman)
David Zafman 2016-05-10 00:30:14 UTC CC vumrao
Flags needinfo?(vumrao)
Vikhyat Umrao 2016-05-10 07:01:04 UTC Flags needinfo?(vumrao) needinfo?(bengland)
David Zafman 2016-05-12 00:29:17 UTC Flags needinfo?(bengland)
Ken Dreyer (Red Hat) 2016-05-18 16:06:51 UTC CC kdreyer
Link ID Ceph Project Bug Tracker 15912
Ken Dreyer (Red Hat) 2016-08-03 19:13:10 UTC Target Release 1.3.3 1.3.4
Jan Buchta 2016-11-14 12:41:35 UTC CC jbuchta
Flags needinfo?(dzafman)
Tomas Petr 2017-02-08 15:04:37 UTC CC tpetr
Tomas Petr 2017-02-08 15:14:59 UTC Flags needinfo?(dzafman)
Vikhyat Umrao 2017-02-08 15:20:32 UTC Target Release 1.3.4 2.3
Vikhyat Umrao 2017-02-08 15:28:52 UTC Blocks 1420417
Jan Buchta 2017-02-10 07:37:43 UTC Flags needinfo?(dzafman)
David Zafman 2017-02-14 18:29:50 UTC Flags needinfo?(dzafman) needinfo?(dzafman) needinfo?(dzafman)
Josh Durgin 2017-04-01 02:24:54 UTC Target Release 2.3 3.0
CC jdurgin
Red Hat Bugzilla Rules Engine 2017-04-01 02:25:00 UTC Target Release 3.0 2.3
Harish NV Rao 2017-04-03 07:09:45 UTC Target Release 2.3 3.0
CC hnallurv
Drew Harris 2017-07-30 15:14:14 UTC Sub Component RADOS
Component Ceph RADOS
Ken Dreyer (Red Hat) 2017-08-15 22:18:37 UTC Flags needinfo?(dzafman)
David Zafman 2017-08-16 00:25:53 UTC Flags needinfo?(dzafman)
Ken Dreyer (Red Hat) 2017-08-16 20:59:13 UTC Status ASSIGNED MODIFIED
Fixed In Version RHEL: ceph-12.1.4-1.el7cp Ubuntu: ceph_12.1.4-2redhat1xenial
errata-xmlrpc 2017-08-16 21:06:49 UTC Status MODIFIED ON_QA
Harish NV Rao 2017-09-16 08:13:31 UTC QA Contact ceph-qe-bugs shmohan
Harish NV Rao 2017-09-28 14:09:46 UTC QA Contact shmohan icolle
Ian Colle 2017-09-29 06:02:25 UTC QA Contact icolle dzafman
Bara Ancincova 2017-10-18 13:01:18 UTC Blocks 1494421
Bara Ancincova 2017-10-24 12:47:54 UTC Docs Contact bancinco
Flags needinfo?(dzafman)
David Zafman 2017-10-24 15:19:44 UTC Doc Text Cause: An OSD disk can become so full that a ceph OSD can't function

Consequence: OSD crashes with a confusing assert message

Fix:

1. Improve error messages
2. By default don't automatically mark more than 25% of OSDs out
3. Improve statfs calculation in filestore/bluestore to better reflect the disk usage

Result: Less likely to run an OSD disk out of space. If it does happen show a better error message in the log.
Flags needinfo?(dzafman)
Bara Ancincova 2017-10-25 09:11:01 UTC Doc Text Cause: An OSD disk can become so full that a ceph OSD can't function

Consequence: OSD crashes with a confusing assert message

Fix:

1. Improve error messages
2. By default don't automatically mark more than 25% of OSDs out
3. Improve statfs calculation in filestore/bluestore to better reflect the disk usage

Result: Less likely to run an OSD disk out of space. If it does happen show a better error message in the log.
.Improvements in handling of full OSDs

When an OSD disk became so full that the OSD could not function, the OSD terminated unexpectedly with a confusing assert message. With this update:

* The error message has been improved.
* By default, more than 25% of OSDs are automatically marked as `out`.
* The `stasfs` calculation in FileStore or BlueStore back ends have been improved to better reflect the disk usage.

As a result, OSDs less likely become full and if this happens, a more informative error message is added to the log.
Flags needinfo?(dzafman)
David Zafman 2017-10-25 17:51:01 UTC Doc Text .Improvements in handling of full OSDs

When an OSD disk became so full that the OSD could not function, the OSD terminated unexpectedly with a confusing assert message. With this update:

* The error message has been improved.
* By default, more than 25% of OSDs are automatically marked as `out`.
* The `stasfs` calculation in FileStore or BlueStore back ends have been improved to better reflect the disk usage.

As a result, OSDs less likely become full and if this happens, a more informative error message is added to the log.
.Improvements in handling of full OSDs

When an OSD disk became so full that the OSD could not function, the OSD terminated unexpectedly with a confusing assert message. With this update:

* The error message has been improved.
* By default, no more than 25% of OSDs are automatically marked as `out`.
* The `statfs` calculation in FileStore or BlueStore back ends have been improved to better reflect the disk usage.

As a result, OSDs are less likely to become full and if they do, a more informative error message is added to the log.
Flags needinfo?(dzafman)
David Zafman 2017-10-30 21:06:01 UTC Status ON_QA VERIFIED
errata-xmlrpc 2017-12-05 17:27:35 UTC Status VERIFIED RELEASE_PENDING
errata-xmlrpc 2017-12-05 23:29:38 UTC Status RELEASE_PENDING CLOSED
Resolution --- ERRATA
Last Closed 2017-12-05 18:29:38 UTC

Back to bug 1332083