Page MenuHomeClusterLabs Projects

Pacemaker 2.0 Tool Changes
Updated 262 Days AgoPublic

Pacemaker 2.0 changed the command-line tools as follows.


Legacy usage (no longer supported)Current usage
-i/--debug_inc and -d/--debug_decn/a (were ignored)
--node-uuid/-un/a (was ignored)
-S/--snmp-traps, -C/--snmp-communityBuilt-in SNMP no longer supported, use alerts instead
-T/--mail-to, -F/--mail-from, -H/--mail-host, -P/--mail-prefixBuilt-in ESMTP no longer supported, use alerts instead
-c/--cman, -A/--openais, -e/--epoch, -H/--heartbeatn/a (legacy stacks no longer supported)
-A/--openais, -H/--heartbeatn/a (legacy stacks no longer supported)
--pending/-jn/a (now always in effect)
tool has been completely removed (was only built with heartbeat)
--ais, --hae, --heartbeat, --lhan/a (legacy stacks no longer supported)
HBDummy test resource agentn/a (heartbeat-class agents no longer supported)
cluster-init (in "extra", not installed by default)
--cman, -h/--heartbeat, -o/--openais, -p/--pluginn/a (legacy stacks no longer supported)
fence agent has been completely removed (was only used with CMAN)
pacemaker plugin for corosync 1
completely removed (was only used with corosync 1)

Reliable exit codes

Prior to 2.0, Pacemaker daemons and command-line tools returned an unreliable mishmash of exit status codes, sometimes including negative numbers (which would get bitwise-remapped to the 0-255 range) and/or C library errno codes (which could potentially vary across OSes).

Now, everything returns a well-defined set of reliable exit status codes. These codes can be viewed using the existing crm_error tool with the --exit parameter. For example:

  • crm_error --exit --list to list all possible exit statuses
  • crm_error --exit 124 to show a textual description of what exit status 124 means

Other changes

  • crm_mon:
    • Output from -X/--as-xml will no longer show expected_votes
  • crm_node:
    • This now always goes through the cluster to obtain information, rather than talk to the cluster layer directly. As a result, the -n/--name, -q/--quorum, and -i/--cluster-id options now work when run on Pacemaker Remote nodes, even if the node name in the cluster is different from local hostname. The -C/--corosync option is deprecated, since the cluster layer interface is no longer needed.
  • crm_report:
    • Data from the heartbeat package is no longer collected
  • crm_resource:
    • The --cleanup option now cleans up only resources with failures; the previous behavior of cleaning up all resources can still be achieved with the --refresh option.
Last Author
Last Edited
Oct 31 2023, 5:58 PM