Pacemaker 2.0.0 - Final

Known regressions introduced in this release

• When run from a Pacemaker Remote node, crm_node might not get the local node name correctly (fixed in 2.0.4)
• Clone notifications can be incorrectly scheduled for a stopped Pacemaker Remote node and block all further cluster actions (fixed in 2.0.1)
• Inccorrect behavior for completing interrupted live migrations (fixed in 2.0.1)
• crm_resource --cleanup could fail to clean up all failures in one run (fixed in 2.0.1)

Deprecated features removed since Pacemaker-1.1.18

• All of these have newer forms, and the cluster will automatically convert most older syntax usage in saved configurations to newer syntax as needed
• Drop support for heartbeat and corosync 1 (whether using CMAN or plugin)
• Drop support for rolling upgrades from Pacemaker versions older than 1.1.11
• Drop support for built-in SMTP and SNMP in crm_mon
• Drop support for legacy option aliases including default-action-timeout, default-resource-stickiness, resource-failure-stickiness,
  default-resource-failure-stickiness, is-managed-default, and all names using underbar instead of dash
• Drop support for "requires" operation meta-attribute
• Drop support for the pcmk_*_cmd, pcmk_arg_map, and pcmk_poweroff_action fence resource parameters
• Drop support for deprecated command-line options to crmadmin, crm_attribute, crm_resource, crm_verify, crm_mon, and stonith_admin
• Drop support for operation meta-attributes in instance_attributes
• Drop support for PCMK_legacy and LRMD_MAX_CHILDREN environment variables
• Drop support for undocumented resource isolation feature
• Drop support for processing very old saved CIB files (including pre-0.6.0 start failure entries, pre-0.6.5 operation history entries,
  pre-0.7 transition keys, pre-1.1.4 migration history entries, pre-1.0 XML configuration schemas, pre-1.1.6 ticket state entries, and
  pre-1.1.7 failed recurring operation history entries)

Features added since Pacemaker-1.1.18

• The pacemaker daemons have been renamed to make logs more intuitive and easier to search
• The default location of the Pacemaker detail log is now /var/log/pacemaker/pacemaker.log (instead of being directly in /var/log), and Pacemaker will no longer use Corosync's logging preferences; configure script options are available to change default log locations
• The detail log's message format has been improved
• The master XML tag is deprecated in favor of using a standard clone tag with a new "promotable" meta-attribute set to true, and the "master-max" and "master-node-max" master meta-attributes are deprecated in favor of new "promoted-max" and "promoted-node-max" clone meta-attributes; documentation now refers to these as promotable clones rather than master/slave, stateful, or multistate clones, and refers to promotion scores instead of master scores
• Administration-related documentation has been moved from the "Pacemaker Explained" document to a new "Pacemaker Administration" document
• record-pending now defaults to TRUE (pending actions are shown in status)
• All Python code in Pacemaker now supports both Python 2.7 and Python 3
• The command-line tools now return consistent, well-defined exit codes; crm_error has an --exit option to list these
• Pacemaker's systemd unit files now remove systemd's spawned process limit
• mount, path, and timer systemd unit types are now supported as resources
• A negative stonith-watchdog-timeout now tells the cluster to automatically calculate the value based on SBD_WATCHDOG_TIMEOUT (which was the behavior of 0 before 1.1.15; 0 retains its post-1.1.15 behavior of disabling use of the watchdog as a fencing device)
• The undocumented restart-type and role_after_failure resource meta-attributes are now deprecated
• Regression testing code has been consolidated and overhauled (the most obvious change is new command names)
• build: create /etc/pacemaker directory when installing
• build: improved portability to BSD-based platforms
• tools: crm_resource --cleanup now cleans only failed operation history; crm_resource --reprobe retains the previous behavior of cleaning all
  operation history
• tools: add stonith_admin --validate option to check device configuration
• tools: crm_node is now in the pacemaker-cli package (instead of pacemaker)
• alerts: add epoch and usec alert variables for improved SNMP alerts
• controller: deprecate "crmd-*" cluster options in favor of new names
• scheduler: deprecate stonith-action value "poweroff" (use "off" instead)
• scheduler: deprecate require-all in rsc_order
• libcrmcluster: prefer corosync name over ring0_addr
• xml: allow local "kind" in resource_set within rsc_order

Changes since Pacemaker-1.1.18

• Restore systemd unit dependency on DBus (regression in 1.1.17)
• CIB: handle mixed-case node names when modifying attributes (regression in 1.1.17)
• scheduler: avoid crash when logging ignored failure timeout (regression in 1.1.17)
• attrd: ensure node name is broadcast at start-up (regression in 1.1.18)
• scheduler: unfence before probing or starting fence devices (regression in 1.1.18)
• tools: treat INFINITY correctly in crm_failcount (regression in 1.1.17)
• tools: show master scores with crm_simulate -sL (regression in 1.1.18)
• tools: crm_master did not work without explicit --lifetime (regression in 1.1.18)
• Numerous changes to public C API of libraries
• Choose current node correctly when a resource is multiply active
• controller,executor,tools: avoid minor memory leaks
• CIB: don't use empty CIB if real CIB has bad permissions
• controller: avoid double free after ACL rejection of resource deletion
• controller: don't record pending clone notifications in CIB
• controller: always write faked failures to CIB whenever possible
• controller: quorum gain without a node join should cause new transition
• executor: handle systemd actions correctly when used with "service:"
• executor: find absolute LSB paths when used with "service:"
• scheduler: handle "requires" of "quorum" or "nothing" properly
• scheduler: ensure orphaned recurring monitors have interval set
• scheduler: handle pending migrations correctly when record-pending is true
• scheduler: don't time out failures that cause fencing until fencing completes
• scheduler: handle globally-unique bundle children correctly
• scheduler: use correct default timeout for monitors
• scheduler: "symmetrical" defaults to "false" for serialize orders
• scheduler: avoid potential use-of-NULL when unpacking ordering constraint
• scheduler: properly cancel recurring monitors
• scheduler: do not schedule notifications for unrunnable actions
• scheduler: ensure stops occur after stopped remote connections come back up
• scheduler: consider only allowed nodes when ordering start after all recovery
• scheduler: avoid graph loop from ordering bundle child stops/demotes after container fencing
• scheduler: remote connection resources are safe to require only quorum
• scheduler: correctly observe colocation with bundles in Master role
• scheduler: restart resource after failed demote when appropriate
• Pacemaker Remote: always use most recent remote proxy
• tools: crm_node now gets correct node name and ID on Pacemaker Remote nodes
• tools: correctly check crm_resource --move for master role
• tools: cibsecret --help/--version doesn't require cluster to be running
• tools: ignore attribute placement when crm_diff compares in cib mode
• tools: prevent notify actions from causing crm_resource --wait to hang
• resources: drop broken configdir parameter from ocf:pacemaker:controld

For further details, see:

• https://wiki.clusterlabs.org/wiki/Pacemaker_2.0_Changes