-
Thu Jan 16 2020 Ken Gaillot <kgaillot@redhat.com> - 1.1.21-4
- Implement shutdown-lock feature
- Resolves: rhbz#1781820
-
Thu Nov 07 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.21-3
- Avoid invalid transition when guest node's host is unclean but can't be fenced
- Resolves: rhbz#1755659
-
Tue Aug 27 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.21-2
- Add latest upstream bug fixes to rebase roll-up patch
- Resolves: rhbz#1731189
-
Tue Jul 30 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.21-1
- Recover from quiesced DC disk
- Avoid timeouts and excessive stonithd CPU usage at start-up in large clusters
- Default serialized order constraints to symmetrical=false
- Avoid fence loops due to incorrect Pacemaker Remote ordering
- Default concurrent-fencing to true
- Harden GnuTLS priorities
- Rebase on upstream 1.1.21 final version
- Resolves: rhbz#1596125
- Resolves: rhbz#1625671
- Resolves: rhbz#1672225
- Resolves: rhbz#1704870
- Resolves: rhbz#1710422
- Resolves: rhbz#1727280
- Resolves: rhbz#1731189
-
Fri May 24 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.20-5
- Correct memory issue in fence agent output fix
- Resolves: rhbz#1549366
-
Fri Apr 19 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.20-4
- Update security patches
- Resolves: rhbz#1694556
- Resolves: rhbz#1694559
- Resolves: rhbz#1694907
-
Thu Apr 04 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.20-3
- Support more than 64KB of fence agent output
- Avoid unnecessary recovery of group member
- Improve IPC clients' authentication of servers (CVE-2018-16877)
- Improve pacemakerd authentication of running subdaemons (CVE-2018-16878)
- Fix use-after-free with potential information disclosure (CVE-2019-3885)
- Resolves: rhbz#1549366
- Resolves: rhbz#1609453
- Resolves: rhbz#1694556
- Resolves: rhbz#1694559
- Resolves: rhbz#1694907
-
Thu Mar 21 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.20-2
- Assume unprivileged ACL if unable to get user information from host
- Delay 2 seconds before re-attempting a failed node attribute write
- SNMP alert sample script now sends all OIDs with every alert
- Recover dependent resources correctly with asymmetric ordering
- Rebase on upstream 1.1.20 final version
- Resolves: rhbz#1596125
- Resolves: rhbz#1597695
- Resolves: rhbz#1608979
- Resolves: rhbz#1628966
- Resolves: rhbz#1644864
-
Fri Feb 01 2019 Ken Gaillot <kgaillot@redhat.com> - 1.1.20-1
- pcs status now shows when a standby node still has active resources
- Allow clean-up of guest nodes and bundles without unmanaging first
- pcs status now shows pending and failed fence actions by default
- Improve pcs status display when disconnected from cluster
- Ensure node attributes are recorded if attrd writer is shutting down
- Synchronize fencing history across all nodes
- Add stonith_admin option to clear fencing history
- Don't schedule unneeded bundle actions when connection is on different node
- Allow use of sbd in clusters with guest nodes and bundles
- Schedule bundle clone notifications correctly when connection is moving
- Rebase on upstream 1.1.20-rc1 version
- Avoid unneeded resource restarts when remote connection fails to start
- Allow crm_resource --move to work when a previous move had a lifetime
- Wait for all replies when refreshing a resource
- Don't schedule clone notifications for a stopped bundle
- Allow option to crm_resource --clear to clear only expired constraints
- Fix result reporting when cleanup is done while an operation is in-flight
- Resolves: rhbz#1419548
- Resolves: rhbz#1448467
- Resolves: rhbz#1461964
- Resolves: rhbz#1486869
- Resolves: rhbz#1535221
- Resolves: rhbz#1555938
- Resolves: rhbz#1595422
- Resolves: rhbz#1627948
- Resolves: rhbz#1638593
- Resolves: rhbz#1644076
- Resolves: rhbz#1644864
- Resolves: rhbz#1648507
- Resolves: rhbz#1648620
- Resolves: rhbz#1652053
- Resolves: rhbz#1652752
- Resolves: rhbz#1658650
- Resolves: rhbz#1665343
-
Mon Sep 24 2018 Ken Gaillot <kgaillot@redhat.com> - 1.1.19-8
- Ensure crm_resource --force-* commands get stderr messages
- Resolves: rhbz#1628947