Skip to content

Latest commit

 

History

History
1255 lines (1110 loc) · 63 KB

CHANGELOG.md

File metadata and controls

1255 lines (1110 loc) · 63 KB
Raw

Change Log

[Unreleased]

Fixed

  • pcs resource manage --monitor no longer enables monitor operation for all resources in a group if only one of the resources was requested to become managed (rhbz#1918527)
  • Misleading error message from pcs booth sync when booth config directory (/etc/booth) is missing (rhbz#1791670)
  • pcs quorum device remove works again (broken since 0.10.13) (rhbz#2115326)

[0.10.14] - 2022-06-23

Security

  • CVE-2022-1049: Pcs daemon was allowing expired accounts, and accounts with expired passwords to login when using PAM auth. (huntr#220307, rhbz#2068456)
  • Pcsd does not expose the server name in HTTP headers anymore (rhbz#2058278)
  • Set Strict-Transport-Security: max-age=63072000 HTTP header for all responses (rhbz#2097392)
  • Set HTTP headers to prevent caching everything except static files (rhbz#2097383)
  • Set HTTP headers to prevent sending referrer (rhbz#2097391)
  • Set cookie option SameSite to Lax (rhbz#2097393)

Added

  • Add support for fence_mpath to pcs stonith update-scsi-devices command (rhbz#2023845)
  • Support for cluster UUIDs. New clusters now get a UUID during setup. Existing clusters can get a UUID by running the new pcs cluster config uuid generate command (rhbz#1950551)
  • Add warning regarding move constraints to pcs status (rhbz#1730232)
  • Support for output formats json and cmd to pcs resource config and pcs stonith config commands (rhbz#1874624, rhbz#1909904)

Fixed

  • Agents not conforming to OCF standard are processed as if they conformed to OCF 1.0 - in the same way as before pcs-0.10.12 (rhbz#2050274)
  • OCF 1.0 agents not conforming to the schema are processed anyway (rhbz#2050274)
  • Booth ticket name validation (rhbz#1791661)
  • Adding booth ticket doesn't report 'mode' as an unknown option anymore (rhbz#1786964)
  • Preventing fence-loop caused when stonith-watchdog-timeout is set with wrong value (rhbz#1954099)
  • Do not allow to create an order constraint for resources in one group as that may block Pacemaker (ghpull#509)

Deprecated

  • Agents not complying with OCF 1.0 schema are processed, incompatibilities are listed as warnings. In pcs-0.11, they will be reported as errors and prevent pcs from working with such agents. (rhbz#2050274)

[0.10.13] - 2022-01-31

Fixed

  • Pcs was not automatically enabling corosync-qdevice when adding a quorum device to a cluster (broken since pcs-0.10.9) (rhbz#2028902)
  • resource update command exiting with a traceback when updating a resource with a non-existing resource agent (rhbz#1384485)
  • pcs_snmp_agent is working again (broken since pcs-0.10.1) (ghpull#431)
  • Skip checking of scsi devices to be removed before unfencing to be added devices (rhbz#2032997)
  • Make ocf:linbit:drbd agent pass OCF standard validation (ghissue#441, rhbz#2036633)
  • Multiple improvements of pcs resource move --autodelete command (rhbz#1990784)
  • Pcs no longer creates Pacemaker-1.x CIB when -f is used, so running pcs cluster cib-upgrade manually is not needed (rhbz#2022463)

[0.10.12] - 2021-11-30

Added

  • Option --autodelete of command pcs resource move is fully supported (rhbz#1990784)
  • Support for OCF 1.1 resource and stonith agents (rhbz#2018969)

Fixed

  • Do not show warning that no stonith device was detected and stonith-enabled is not false when a stonith device is in a group (ghpull#370)
  • Misleading error message from pcs quorum unblock when wait_for_all=0 (rhbz#1968088)
  • Misleading error message from pcs booth setup and pcs booth pull when booth config directory (/etc/booth) is missing (rhbz#1791670, ghpull#411, ghissue#225)

[0.10.11] - 2021-10-05

Added

  • Add add/remove cli syntax for command pcs stonith update-scsi-devices (rhbz#1992668)

Fixed

  • Fixed an error when creating a resource which defines 'depth' attribute for its operations (rhbz#1998454)
  • Do not unfence newly added devices on fenced cluster nodes (rhbz#1991654)
  • Fix displaying fencing levels with regular expression targets (rhbz#1533090)

[0.10.10] - 2021-08-19

Added

  • Support for new role names introduced in pacemaker 2.1 (rhbz#1885293)

Fixed

  • Traceback in some cases when --wait without timeout is used

[0.10.9] - 2021-08-10

Added

  • Elliptic curve TLS certificates are now supported in pcsd (ghissue#123)
  • Support for corosync option totem.block_unlisted_ips (rhbz#1720221)
  • Support for displaying status of a single resource or tag (rhbz#1290830)
  • Support for displaying status of resources on a specified node (rhbz#1285269)
  • New option --brief for pcs resource disable --safe or its alias pcs resource safe-disable that only prints errors (rhbz#1909901)
  • Support for updating scsi fencing devices without affecting other resources added in the new command pcs stonith update-scsi-devices (rhbz#1759995, rhbz#1872378)
  • Option --autodelete for pcs resource move command which removes a location constraint used for moving a resource, once the resource has been moved. This feature is in tech-preview state and thus may be changed in the future (rhbz#1847102)

Fixed

  • Node attribute expressions are now correctly reported as not allowed in resource defaults rules (rhbz#1896458)
  • Upgreded to jquery 3.6.0 ([rhbz#1882291, rhbz#1886342])
  • Man page and help: note that 'pcs resource unclone' accepts clone resources as well (rhbz#1930886)
  • Improved error messages when a host is found to be a part of a cluster already (rhbz#1690419)
  • pcs cluster sync command now warns reloading corosync config is necessary for changes to take effect (rhbz#1750240)
  • Show user friendly error if unable to delete a group (due to the group being referenced within configuration) when moving resources out of the the group. (rhbz#1678273)
  • Exit with an error if on-fail=demote is specified for a resource operation and pacemaker doesn't support it
  • The pcs status nodes command now correctly shows status of nodes that are both in maintenance and standby modes (rhbz#1432097)

Changed

  • python3-openssl was replaced with python3-cryptography (rhbz#1927404)

Deprecated

  • pcs acl show replaced with pcs acl config
  • pcs alert show replaced with pcs alert config
  • Undocumented command pcs cluster certkey replaced with pcs pcsd certkey
  • pcs cluster pcsd-status replaced with pcs status pcsd or pcs pcsd status
  • pcs constraint [location | colocation | order | ticket] show | list replaced with pcs constraint [location | colocation | order | ticket] config
  • pcs property show, pcs property list replaced with pcs property config
  • pcsd urls: /remote/config_backup, /remote/node_available, /remote/node_restart, /remote/resource_status
  • Undocumented syntax for constraint location rules:
    • date start=<date> gt replaced with date gt <date>
    • date end=<date> lt replaced with date lt <date>
    • date start=<date> end=<date> in_range replaced with date in_range <date> to <date>
    • operation=date_spec replaced with date-spec <date-spec options>
    • converting invalid score to score-attribute=pingd
  • Delimiting stonith devices with a comma in pcs stonith level add | clear | delete | remove commands, use a space instead
  • pcs stonith level delete | remove [<target>] [<stonith id>]... replaced with pcs stonith level delete | remove [target <target>] [stonith <stonith id>]...
  • pcs stonith level clear [<target> | <stonith ids>] replaced with pcs stonith level clear [target <target> | stonith <stonith id>...]
  • pcs tag list replaced with pcs tag config

[0.10.8] - 2021-02-01

Added

Fixed

  • Improved error message with a hint in pcs cluster cib-push (ghissue#241)
  • Option --wait was not working with pacemaker 2.0.5+ (ghissue#260)
  • Explicitly close libcurl connections to prevent stalled TCP connections in CLOSE-WAIT state (ghissue#261, rhbz#1885841)
  • Fixed parsing negative float numbers on command line (rhbz#1869399)
  • Removed unwanted logging to system log (/var/log/messages) (rhbz#1917286)
  • Fixed rare race condition in pcs cluster start --wait (rhbz#1794062)
  • Better error message when unable to connect to pcsd (rhbz#1619818)

Deprecated

  • Commands pcs config import-cman and pcs config export pcs-commands|pcs-commands-verbose have been deprecated (rhbz#1851335)
  • Entering values starting with '-' (negative numbers) without '--' on command line is now deprecated (rhbz#1869399)

[0.10.7] - 2020-09-30

Added

Fixed

  • Prevent removing non-empty tag by removing tagged resource group or clone (rhbz#1857295)
  • Clarify documentation for 'resource move' and 'resource ban' commands with regards to the 'lifetime' option.
  • Allow moving both promoted and demoted promotable clone resources (rhbz#1875301)

Deprecated

  • pcs resource [op] defaults <name>=<value>... commands are deprecated now. Use pcs resource [op] defaults update <name>=<value>... if you only manage one set of defaults, or pcs resource [op] defaults set if you manage several sets of defaults. (rhbz#1817547)

[0.10.6] - 2020-06-11

Security

  • Web UI sends HTTP headers: Content-Security-Policy, X-Frame-Options and X-Xss-Protection

Added

  • When creating a cluster, verify the cluster name does not prevent mounting GFS2 volumes (rhbz#1782553)
  • An option to run 'pcs cluster setup' in a local mode (do not connect to any nodes, save corosync.conf to a specified file) (rhbz#1839637)
  • Support for pacemaker tags. Pcs provides commands for creating and removing tags, adding and/or removing IDs to/from tags, and listing current tag configuration. (rhbz#1684676)
  • Support for tag ids in commands resource enable/disable/manage/unmanage (rhbz#1684676)
  • pcs resource [safe-]disable --simulate has a new option --brief to print only a list of affected resources (rhbz#1833114)

Fixed

  • Keep autogenerated IDs of set constraints reasonably short (rhbz#1387358, rhbz#1824206)
  • Pcs is now compatible with Ruby 2.7 and Python 3.8. To achieve this, it newly depends on python3-distro package.
  • pcs status works on remote nodes again (broken since pcs-0.10.4) (rhbz#1830552)
  • Fixed inability to create colocation constraint from web ui (rhbz#1832973)
  • Actions going through pcsd no longer time out after 30s (broken since pcs-0.10.5) (rhbz#1833506)

[0.10.5] - 2020-03-18

Added

  • It is possible to configure a disaster-recovery site and display its status (rhbz#1676431)

Fixed

  • Error messages in cases when cluster is not set up (rhbz#1743731)
  • Improved documentation of configuring links in the 'pcs cluster setup' command
  • Safe-disabling clones and groups does not fail any more due to their inner resources get stopped (rhbz#1781303)
  • Booth documentation clarified (ghissue#231)
  • Detection of fence history support (rhbz#1793574)
  • Fix documentation and flags regarding bundled/cloned/grouped resources for pcs (resource | stonith) (cleanup | refresh) (rhbz#1805082)
  • Improved ACL documentation (rhbz#1722970)
  • Added missing Strict-Transport-Security headers to redirects (rhbz#1810017)
  • Improved pcsd daemon performance (rhbz#1783106)

[0.10.4] - 2019-11-28

Added

  • New section in pcs man page summarizing changes in pcs-0.10. Commands removed or changed in pcs-0.10 print errors poiting to that section. (rhbz#1728890)
  • pcs resource disable can show effects of disabling resources and prevent disabling resources if any other resources would be affected (rhbz#1631519)
  • pcs resource relations command shows relations between resources such as ordering constraints, ordering set constraints and relations defined by resource hierarchy (rhbz#1631514)

Changed

  • Expired location constraints are now hidden by default when listing constraints in any way. Using --all will list and denote them with (expired). All expired rules are then marked the same way. (rhbz#1442116)

Fixed

  • All node names and scores are validated when running pcs constraint location avoids/prefers before writing configuration to cib (rhbz#1673835)
  • Fixed crash when an invalid port is given in an address to the pcs host auth command (rhbz#1698763)
  • Command pcs cluster verify suggests --full option instead of -V option which is not recognized by pcs (rhbz#1712347)
  • It is now possible to authenticate remote clusters in web UI even if the local cluster is not authenticated (rhbz#1743735)
  • Documentation of pcs constraint colocation add (rhbz#1734361)
  • Empty constraint option are not allowed in pcs constraint order and pcs constraint colocation add commands (rhbz#1734361)
  • More fixes for the case when PATH environment variable is not set
  • Fixed crashes and other issues when UTF-8 characters are present in the corosync.conf file (rhbz#1741586)

[0.10.3] - 2019-08-23

Fixed

  • Fixed crashes in the pcs host auth command (rhbz#1676957)
  • Fixed id conflict with current bundle configuration in pcs resource bundle reset (rhbz#1657166)
  • Options starting with - and -- are no longer ignored for non-root users (broken since pcs-0.10.2) (rhbz#1725183)
  • Fixed crashes when pcs is configured that no rubygems are bundled in pcs package (ghissue#208)
  • Standby nodes running resources are listed separately in pcs status nodes
  • Parsing arguments in the pcs constraint order and pcs constraint colocation add commands has been improved, errors which were previously silent are now reported (rhbz#1734361)
  • Fixed shebang correction in Makefile (ghissue#206)
  • Generate 256 bytes long corosync authkey, longer keys are not supported when FIPS is enabled (rhbz#1740218)

Changed

  • Command pcs resource bundle reset no longer accepts the container type (rhbz#1657166)

[0.10.2] - 2019-06-12

Added

  • Command pcs config checkpoint diff for displaying differences between two specified checkpoints (rhbz#1655055)
  • Support for resource instance attributes uniqueness check according to resource agent metadata (rhbz#1665404)
  • Command pcs resource bundle reset for a bundle configuration resetting (rhbz#1657166)
  • pcs cluster setup now checks if nodes' addresses match value of ip_version (rhbz#1667053)
  • Support for sbd option SBD_TIMEOUT_ACTION (rhbz#1664828)
  • Support for clearing expired moves and bans of resources (rhbz#1625386)
  • Commands for adding, changing and removing corosync links (rhbz#1667058)

Fixed

  • Corosync config file parser updated and made more strict to match changes in corosync
  • Allow non-root users to read quorum status (commands pcs status corosync, pcs status quorum, pcs quorum device status, pcs quorum status) (rhbz#1653316)
  • Removed command pcs resource show dropped from usage and man page (rhbz#1656953)
  • Put proper link options' names to corosync.conf (rhbz#1659051)
  • Fixed issuses in configuring links in the 'create cluster' form in web UI (rhbz#1664057)
  • Pcs no longer removes empty meta_attributes, instance_attributes and other nvsets and similar elements from CIB. Such behavior was causing problems when pacemaker ACLs were in effect, leading to inability of pushing modified CIBs to pacemaker. (rhbz#1659144)
  • ipv4-6 and ipv6-4 are now valid values of ip_version in cluster setup (rhbz#1667040)
  • Crash when using unsupported options in commands pcs status and pcs config (rhbz#1668422)
  • pcs resource group add now fails gracefully instead of dumping an invalid CIB when a group ID is already occupied by a non-resource element (rhbz#1668223)
  • pcs no longer spawns unnecessary processes for reading known hosts (rhbz#1676945)
  • Lower load caused by periodical config files syncing in pcsd by making it sync less frequently (rhbz#1676957)
  • Improve logging of periodical config files syncing in pcsd
  • Knet link option ip_version has been removed, it was never supported by corosync. Transport option ip_version is still in place. (rhbz#1674005)
  • Several bugs in linklist validation in pcs cluster setup (rhbz#1667090)
  • Fixed a typo in documentation (regardles -> regardless) (rhbz#1660702)
  • Fixed pcsd crashes when non-ASCII characters are present in systemd journal
  • Pcs works even when PATH environment variable is not set (rhbz#1673825)
  • Fixed several "Unknown report" error messages
  • Pcsd SSL certificates are no longer synced across cluster nodes when creating new cluster or adding new node to an existing cluster. To enable the syncing, set PCSD_SSL_CERT_SYNC_ENABLED to true in pcsd config. (rhbz#1673822)
  • Pcs now reports missing node names in corosync.conf instead of failing silently
  • Fixed an issue where some pcs commands could not connect to cluster nodes over IPv6
  • Fixed cluster setup problem in web UI when full domain names are used (rhbz#1687965)
  • Fixed inability to setup cluster in web UI when knet links are not specified (rhbz#1687562)
  • --force works correctly in pcs quorum unblock (broken since pcs-0.10.1)
  • Removed 3des from allowed knet crypto ciphers since it is actually not supported by corosync
  • Improved validation of corosync options and their values (rhbz#1679196, rhbz#1679197)

Changed

  • Do not check whether watchdog is defined as an absolute path when enabling SBD. This check is not needed anymore as we are validating watchdog against list provided by SBD itself.

Deprecated

  • Command pcs resource show, removed in pcs-0.10.1, has been readded as deprecated to ease transition to its replacements. It will be removed again in future. rhbz#1661059

[0.10.1] - 2018-11-23

Removed

  • Pcs-0.10 removes support for CMAN, Corosync 1.x, Corosync 2.x and Pacemaker 1.x based clusters. For managing those clusters use pcs-0.9.x.
  • Pcs-0.10 requires Python 3.6 and Ruby 2.2, support for older Python and Ruby versions has been removed.
  • pcs resource failcount reset command has been removed as pcs resource cleanup is doing exactly the same job. (rhbz#1427273)
  • Deprecated commands pcs cluster remote-node add | remove have been removed as they were replaced with pcs cluster node add-guest | remove-guest
  • Ability to create master resources has been removed as they are deprecated in Pacemaker 2.x (rhbz#1542288)
    • Instead of pcs resource create ... master use pcs resource create ... promotable or pcs resource create ... clone promotable=true
    • Instead of pcs resource master use pcs resource promotable or pcs resource clone ... promotable=true
  • Deprecated --clone option from pcs resource create command
  • Ability to manage node attributes with pcs property set|unset|show commands (using --node option). The same functionality is still available using pcs node attribute command.
  • Undocumented version of the pcs constraint colocation add command, its syntax was pcs constraint colocation add <source resource id> <target resource id> [score] [options]
  • Deprecated commands pcs cluster standby | unstandby, use pcs node standby | unstandby instead
  • Deprecated command pcs cluster quorum unblock which was replaced by pcs quorum unblock
  • Subcommand pcs status groups as it was not showing a cluster status but cluster configuration. The same functionality is still available using command pcs resource group list
  • Undocumented command pcs acl target, use pcs acl user instead

Added

  • Validation for an unaccessible resource inside a bundle (rhbz#1462248)
  • Options to filter failures by an operation and its interval in pcs resource cleanup and pcs resource failcount show commands (rhbz#1427273)
  • Commands for listing and testing watchdog devices (rhbz#1578891)
  • Commands for creating promotable clone resources pcs resource promotable and pcs resource create ... promotable (rhbz#1542288)
  • pcs resource update and pcs resource meta commands change master resources to promotable clone resources because master resources are deprecated in Pacemaker 2.x (rhbz#1542288)
  • Support for the promoted-max bundle option replacing the masters option in Pacemaker 2.x (rhbz#1542288)
  • Support for OP_NO_RENEGOTIATION option when OpenSSL supports it (even with Python 3.6) (rhbz#1566430)
  • Support for container types rkt and podman into bundle commands (rhbz#1619620)
  • Support for promotable clone resources in pcsd and web UI (rhbz#1542288)
  • Obsoleting parameters of resource and fence agents are now supported and preferred over deprecated parameters (rhbz#1436217)
  • pcs status now shows failed and pending fencing actions and pcs status --full shows the whole fencing history. Pacemaker supporting fencing history is required. (rhbz#1615891)
  • pcs stonith history commands for displaying, synchronizing and cleaning up fencing history. Pacemaker supporting fencing history is required. (rhbz#1620190)
  • Validation of node existence in a cluster when creating location constraints (rhbz#1553718)
  • Command pcs client local-auth for authentication of pcs client against local pcsd. This is required when a non-root user wants to execute a command which requires root permissions (e.g. pcs cluster start). (rhbz#1554302)
  • Command pcs resource group list which has the same functionality as removed command pcs resource show --groups

Fixed

  • Fixed encoding of the CIB_user_groups cookie in communication between nodes.
  • pcs cluster cib-push diff-against= does not consider an empty diff as an error (ghpull#166)
  • pcs cluster cib-push diff-against= exits gracefully with an error message if crm_feature_set < 3.0.9 (rhbz#1488044)
  • pcs resource update does not create an empty meta_attributes element any more (rhbz#1568353)
  • pcs resource debug-* commands provide debug messages even with pacemaker-1.1.18 and newer (rhbz#1574898)
  • Improve pcs quorum device add usage and man page (rhbz#1476862)
  • Removing resources using web UI when the operation takes longer than expected (rhbz#1579911)
  • Removing a cluster node no longer leaves the node in the CIB and therefore cluster status even if the removal is run on the node which is being removed (rhbz#1595829)
  • Possible race condition causing an HTTP 408 error when sending larger files via pcs (rhbz#1600169)
  • Configuring QDevice works even if NSS with the new db format (cert9.db, key4.db, pkcs11.txt) is used (rhbz#1596721)
  • Options starting with '-' and '--' are no longer accepted by commands for which those options have no effect (rhbz#1533866)
  • When a user makes an error in a pcs command, usage for that specific command is printed instead of printing the whole usage
  • Show more user friendly error message when testing watchdog device and multiple devices are present (rhbz#1578891)
  • Do not distinguish between supported and unsupported watchdog devices as SBD cannot reliably provide such information (rhbz#1578891)
  • pcs config no longer crashes when crm_mon prints something to stderr (rhbz#1578955)
  • pcs resource bundle update cmd for bundles which are using unsupported container backend (rhbz#1619620)
  • Do not crash if unable to load SSL certificate or key, log errors and exit gracefully instead (rhbz#1638852)
  • Fixed several issues in parsing pcs constraint colocation add command.
  • All remove subcommands now have delete aliases and vice versa. Previously, only some of them did and it was mostly undocumented.
  • The pcs acl role delete command no longer deletes ACL users and groups with no ACL roles assigned

Changed

  • Authentication has been overhauled (rhbz#1549535):
    • The pcs cluster auth command only authenticates nodes in a local cluster and does not accept a node list.
    • The new command for authentication is pcs host auth. It allows to specify host names, addresses and pcsd ports.
    • Previously, running pcs cluster auth A B C caused A, B and C to be all authenticated against each other. Now, pcs host auth A B C makes the local host authenticated against A, B and C. This allows better control of what is authenticated against what.
    • The pcs pcsd clear-auth command has been replaced by pcs pcsd deauth and pcs host deauth commands. The new commands allows to deauthenticate a single host / token as well as all hosts / tokens.
    • These changes are not backward compatible. You should use the pcs host auth command to re-authenticate your hosts.
  • The pcs cluster setup command has been overhauled (rhbz#1158816, rhbz#1183103):
    • It works with Corosync 3.x only and supports knet as well as udp/udpu.
    • Node names are now supported.
    • The number of Corosync options configurable by the command has been significantly increased.
    • The syntax of the command has been completely changed to accommodate the changes and new features.
    • Corosync encryption is enabled by default when knet is used (rhbz#1648942)
  • The pcs cluster node add command has been overhauled (rhbz#1158816, rhbz#1183103)
    • It works with Corosync 3.x only and supports knet as well as udp/udpu.
    • Node names are now supported.
    • The syntax of the command has been changed to accommodate new features and to be consistent with other pcs commands.
  • The pcs cluster node remove has been overhauled (rhbz#1158816, rhbz#1595829):
    • It works with Corosync 3.x only and supports knet as well as udp/udpu.
    • It is now possible to remove more than one node at once.
    • Removing a cluster node no longer leaves the node in the CIB and therefore cluster status even if the removal is run on the node which is being removed
  • Node names are fully supported now and are no longer coupled with node addresses. It is possible to set up a cluster where Corosync communicates over different addresses than pcs/pcsd. (rhbz#1158816, rhbz#1183103)
  • Node names are now required while node addresses are optional in the pcs cluster node add-guest and pcs cluster node add-remove commands. Previously, it was the other way around.
  • Web UI has been updated following changes in authentication and support for Corosync 3.x (rhbz#1158816, rhbz#1183103, rhbz#1549535)
  • Commands related to resource failures have been overhauled to support changes in pacemaker. Failures are now tracked per resource operations on top of resources and nodes. (rhbz#1427273, rhbz#1588667)
  • --watchdog and --device options of pcs stonith sbd enable and pcs stonith sbd device setup commands have been replaced with watchdog and device options respectively
  • Update pacemaker daemon names to match changes in pacemaker-2.0 (rhbz#1573344)
  • Watchdog devices are validated against a list provided by sbd (rhbz#1578891)
  • Resource operation option requires is no longer accepted to match changes in pacemaker-2.0 (rhbz#1605185)
  • Update pacemaker exit codes to match changes in pacemaker-2.0 (rhbz#1536121)
  • pcs cluster cib-upgrade no longer exits with an error if the CIB schema is already the latest available (this has been changed in pacemaker-2.0)
  • Pcs now configures corosync to put timestamps in its log (rhbz#1615420)
  • Option -V has been replaced with --full and a CIB file can be specified only using option -f in pcs cluster verify
  • Master resources are now called promotable clone resources to match changes in pacemaker-2.0 (rhbz#1542288)
  • Key size of default pcsd self-generated certificates increased from 2048b to 3072b (rhbz#1638852)
  • pcsd.service now depends on network-online.target (rhbz#1640477)
  • Split command pcs resource [show] into two new commands:
    • pcs resource [status] - same as pcs resource [show]
    • pcs resource config - same as pcs resource [show] --full or resource id specified instead of --full Respective changes have been made to pcs stonith [show] command.
  • Previously, pcs cluster sync synchronized only corosync configuration across all nodes configured in the cluster. This command will be changed in the future to sync all cluster configuration. New subcommand pcs cluster sync corosync has been introduced to sync only corosync configuration. For now, both commands have the same functionality.

Security

  • CVE-2018-1086: Debug parameter removal bypass, allowing information disclosure (rhbz#1557366)
  • CVE-2018-1079: Privilege escalation via authorized user malicious REST call (rhbz#1550243)

Deprecated

  • The masters bundle option is obsoleted by the promoted-max option in Pacemaker 2.x and therefore in pcs (rhbz#1542288)
  • pcs cluster uidgid rm, use pcs cluster uidgid delete or pcs cluster uidgid remove instead

[0.9.163] - 2018-02-20

Added

  • Added pcs status booth as an alias to pcs booth status
  • A warning is displayed in pcs status and a stonith device detail in web UI when a stonith device has its method option set to cycle (rhbz#1523378)

Fixed

  • --skip-offline is no longer ignored in the pcs quorum device remove command
  • pcs now waits up to 5 minutes (previously 10 seconds) for pcsd restart when synchronizing pcsd certificates
  • Usage and man page now correctly state it is possible to enable or disable several stonith devices at once
  • It is now possible to set the action option of stonith devices in web UI by using force (rhbz#1421702)
  • Do not crash when --wait is used in pcs stonith create (rhbz#1522813)
  • Nodes are now authenticated after running pcs cluster auth even if an existing corosync.conf defines no nodes (ghissue#153, rhbz#1517333)
  • Pcs now properly exits with code 1 when an error occurs in pcs cluster node add-remote and pcs cluster node add-guest commands (rhbz#1464781)
  • Fixed a crash in the pcs booth sync command (rhbz#1527530)
  • Always replace the whole CIB instead of applying a diff when crm_feature_set <= 3.0.8 (rhbz#1488044)
  • Fixed pcs cluster auth in a cluster when not authenticated and using a non-default port (rhbz#1415197)
  • Fixed pcs cluster auth in a cluster when previously authenticated using a non-default port and reauthenticating using an implicit default port (rhbz#1415197)

[0.9.162] - 2017-11-15

Added

  • pcs status --full now displays information about tickets (rhbz#1389943)
  • Support for managing qdevice heuristics (rhbz#1389209)
  • SNMP agent providing information about cluster to the master agent. It supports only python 2.7 for now (rhbz#1367808).

Fixed

  • Fixed crash when loading a huge xml (rhbz#1506864)
  • Fixed adding an existing cluster into the web UI (rhbz#1415197)
  • False warnings about failed actions when resource is master/unmaster from the web UI (rhbz#1506220)

Changed

  • pcs resource|stonith cleanup no longer deletes the whole operation history of resources. Instead, it only deletes failed operations from the history. The original functionality is available in the pcs resource|stonith refresh command. (rhbz#1508351, rhbz#1508350)

[0.9.161] - 2017-11-02

Added

Fixed

  • Fixed pcs cluster auth when already authenticated and using different port (rhbz#1415197)
  • It is now possible to restart a bundle resource on one node (rhbz#1501274)
  • resource update no longer exits with an error when the remote-node meta attribute is set to the same value that it already has (rhbz#1502715, ghissue#145)
  • Listing and describing resource and stonith agents no longer crashes when agents' metadata contain non-ascii characters (rhbz#1503110, ghissue#151)

[0.9.160] - 2017-10-09

Added

Fixed

  • Fixed some crashes when pcs encounters a non-ascii character in environment variables, command line arguments and so on (rhbz#1435697)
  • Fixed detecting if systemd is in use (ghissue#118)
  • Upgrade CIB schema version when resource-discovery option is used in location constraints (rhbz#1420437)
  • Fixed error messages in pcs cluster report (rhbz#1388783)
  • Increase request timeout when starting a cluster with large number of nodes to prevent timeouts (rhbz#1463327)
  • Fixed "Unable to update cib" error caused by invalid resource operation IDs
  • pcs resource op defaults now fails on an invalid option (rhbz#1341582)
  • Fixed behaviour of pcs cluster verify command when entered with the filename argument (rhbz#1213946)

Changed

  • CIB changes are now pushed to pacemaker as a diff in commands overhauled to the new architecture (previously the whole CIB was pushed). This resolves race conditions and ACLs related errors when pushing CIB. (rhbz#1441673)
  • All actions / operations defined in resource agent's metadata (except meta-data, status and validate-all) are now copied to the CIB when creating a resource. (rhbz#1418199, ghissue#132)
  • Improve documentation of the pcs stonith confirm command (rhbz#1489682)

Deprecated

  • This is the last version fully supporting CMAN clusters and python 2.6. Support for these will be gradually dropped.

[0.9.159] - 2017-06-30

Added

  • Option to create a cluster with or without corosync encryption enabled, by default the encryption is disabled (rhbz#1165821)
  • It is now possible to disable, enable, unmanage and manage bundle resources and set their meta attributes (rhbz#1447910)
  • Pcs now warns against using the action option of stonith devices (rhbz#1421702)

Fixed

  • Fixed crash of the pcs cluster setup command when the --force flag was used (rhbz#1176018)
  • Fixed crash of the pcs cluster destroy --all command when the cluster was not running (rhbz#1176018)
  • Fixed crash of the pcs config restore command when restoring pacemaker authkey (rhbz#1176018)
  • Fixed "Error: unable to get cib" when adding a node to a stopped cluster (rhbz#1176018)
  • Fixed a crash in the pcs cluster node add-remote command when an id conflict occurs (rhbz#1386114)
  • Fixed creating a new cluster from the web UI (rhbz#1284404)
  • pcs cluster node add-guest now works with the flag --skip-offline (rhbz#1176018)
  • pcs cluster node remove-guest can be run again when the guest node was unreachable first time (rhbz#1176018)
  • Fixed "Error: Unable to read /etc/corosync/corosync.conf" when running pcs resource create(rhbz#1386114)
  • It is now possible to set debug and verbose parameters of stonith devices (rhbz#1432283)
  • Resource operation ids are now properly validated and no longer ignored in pcs resource create, pcs resource update and pcs resource op add commands (rhbz#1443418)
  • Flag --force works correctly when an operation is not successful on some nodes during pcs cluster node add-remote or pcs cluster node add-guest (rhbz#1464781)

Changed

  • Binary data are stored in corosync authkey (rhbz#1165821)
  • It is now mandatory to specify container type in the resource bundle create command
  • When creating a new cluster, corosync communication encryption is disabled by default (in 0.9.158 it was enabled by default, in 0.9.157 and older it was disabled)

[0.9.158] - 2017-05-23

Added

  • Support for bundle resources (CLI only) (rhbz#1433016)
  • Commands for adding and removing guest and remote nodes including handling pacemaker authkey (CLI only) (rhbz#1176018, rhbz#1254984, rhbz#1386114, rhbz#1386512)
  • Command pcs cluster node clear to remove a node from pacemaker's configuration and caches
  • Backing up and restoring cluster configuration by pcs config backup and pcs config restore commands now support corosync and pacemaker authkeys (rhbz#1165821, rhbz#1176018)

Deprecated

  • pcs cluster remote-node add and pcs cluster remote-node remove commands have been deprecated in favor of pcs cluster node add-guest and pcs cluster node remove-guest commands (rhbz#1386512)

Fixed

  • Fixed a bug which under specific conditions caused pcsd to crash on start when running under systemd (ghissue#134)
  • pcs resource unmanage now sets the unmanaged flag to primitive resources even if a clone or master/slave resource is specified. Thus the primitive resources will not become managed just by uncloning. This also prevents some discrepancies between disabled monitor operations and the unmanaged flag. (rhbz#1303969)
  • pcs resource unmanage --monitor now properly disables monitor operations even if a clone or master/slave resource is specified. (rhbz#1303969)
  • --help option now shows help just for the specified command. Previously the usage for a whole group of commands was shown.
  • Fixed a crash when pcs cluster cib-push is called with an explicit value of the --wait flag (rhbz#1422667)
  • Handle pcsd crash when an unusable address is set in PCSD_BIND_ADDR (rhbz#1373614)
  • Removal of a pacemaker remote resource no longer causes the respective remote node to be fenced (rhbz#1390609)

Changed

[0.9.157] - 2017-04-10

Added

  • Resources in location constraints now may be specified by resource name patterns in addition to resource names (rhbz#1362493)
  • Proxy settings description in pcsd configuration file (rhbz#1315627)
  • Man page for pcsd (rhbz#1378742)
  • Pcs now allows to set trace_ra and trace_file options of ocf:heartbeat and ocf:pacemaker resources (rhbz#1421702)
  • pcs resource describe and pcs stonith describe commands now show all information about the specified agent if the --full flag is used
  • pcs resource manage | unmanage enables respectively disables monitor operations when the --monitor flag is specified (rhbz#1303969)
  • Support for shared storage in SBD. Currently, there is very limited support in web UI (rhbz#1413958)

Changed

  • It is now possible to specify more than one resource in the pcs resource enable and pcs resource disable commands.

Fixed

  • Python 3: pcs no longer spams stderr with error messages when communicating with another node
  • Stopping a cluster does not timeout too early and it generally works better even if the cluster is running Virtual IP resources (rhbz#1334429)
  • pcs booth remove now works correctly even if the booth resource group is disabled (another fix) (rhbz#1389941)
  • Fixed Cross-site scripting (XSS) vulnerability in web UI (CVE-2017-2661, rhbz#1434111)
  • Pcs no longer allows to create a stonith resource based on an agent whose name contains a colon (rhbz#1415080)
  • Pcs command now launches Python interpreter with "sane" options (python -Es) (rhbz#1328882)
  • Clufter is now supported on both Python 2 and Python 3 (rhbz#1428350)
  • Do not colorize clufter output if saved to a file

[0.9.156] - 2017-02-10

Added

  • Fencing levels now may be targeted in CLI by a node name pattern or a node attribute in addition to a node name (rhbz#1261116)
  • pcs cluster cib-push allows to push a diff obtained internally by comparing CIBs in specified files (rhbz#1404233, rhbz#1419903)
  • Added flags --wait, --disabled, --group, --after, --before into the command pcs stonith create
  • Added commands pcs stonith enable and pcs stonith disable
  • Command line option --request-timeout (rhbz#1292858)
  • Check whenever proxy is set when unable to connect to a node (rhbz#1315627)

Changed

  • pcs node [un]standby and pcs node [un]maintenance is now atomic even if more than one node is specified (rhbz#1315992)
  • Restarting pcsd initiated from pcs is now a synchronous operation (rhbz#1284404)
  • Stopped bundling fonts used in pcsd web UI (ghissue#125)
  • In pcs resource create flags --master and --clone changed to keywords master and clone
  • libcurl is now used for node to node communication

Fixed

  • When upgrading CIB to the latest schema version, check for minimal common version across the cluster (rhbz#1389443)
  • pcs booth remove now works correctly even if the booth resource group is disabled (rhbz#1389941)
  • Adding a node in a CMAN cluster does not cause the new node to be fenced immediately (rhbz#1394846)
  • Show proper error message when there is an HTTP communication failure (rhbz#1394273)
  • Fixed searching for files to remove in the /var/lib directory (ghpull#119, ghpull#120)
  • Fixed messages when managing services (start, stop, enable, disable...)
  • Fixed disabling services on systemd systems when using instances (rhbz#1389501)
  • Fixed parsing commandline options (rhbz#1404229)
  • Pcs does not exit with a false error message anymore when pcsd-cli.rb outputs to stderr (ghissue#124)
  • Pcs now exits with an error when both --all and a list of nodes is specified in the pcs cluster start | stop | enable | disable commands (rhbz#1339355)
  • built-in help and man page fixes and improvements (rhbz#1347335)
  • In pcs resource create the flag --clone no longer steals arguments from the keywords meta and op (rhbz#1395226)
  • pcs resource create does not produce invalid cib when group id is already occupied with non-resource element (rhbz#1382004)
  • Fixed misbehavior of the flag --master in pcs resource create command (rhbz#1378107)
  • Fixed tacit acceptance of invalid resource operation in pcs resource create (rhbz#1398562)
  • Fixed misplacing metadata for disabling when running pcs resource create with flags --clone and --disabled (rhbz#1402475)
  • Fixed incorrect acceptance of the invalid attribute of resource operation in pcs resource create (rhbz#1382597)
  • Fixed validation of options of resource operations in pcs resource create (rhbz#1390071)
  • Fixed silent omission of duplicate options (rhbz#1390066)
  • Added more validation for resource agent names (rhbz#1387670)
  • Fixed network communication issues in pcsd when a node was specified by an IPv6 address
  • Fixed JS error in web UI when empty cluster status is received (rhbz#1396462)
  • Fixed sending user group in cookies from Python 3
  • Fixed pcsd restart in Python 3
  • Fixed parsing XML in Python 3 (caused crashes when reading resource agents metadata) (rhbz#1419639)
  • Fixed the recognition of the structure of a resource agent name that contains a systemd instance (rhbz#1419661)

Removed

  • Ruby 1.8 and 1.9 is no longer supported due to bad libcurl support

[0.9.155] - 2016-11-03

Added

  • Show daemon status in pcs status on non-systemd machines
  • SBD support for cman clusters (rhbz#1380352)
  • Alerts management in pcsd (rhbz#1376480)

Changed

  • Get all information about resource and stonith agents from pacemaker. Pcs now supports the same set of agents as pacemaker does. (rhbz#1262001, ghissue#81)
  • pcs resource create now exits with an error if more than one resource agent matches the specified short agent name instead of randomly selecting one of the agents
  • Allow to remove multiple alerts and alert recipients at once

Fixed

  • When stopping a cluster with some of the nodes unreachable, stop the cluster completely on all reachable nodes (rhbz#1380372)
  • Fixed pcsd crash when rpam rubygem is installed (ghissue#109)
  • Fixed occasional crashes / failures when using locale other than en_US.UTF8 (rhbz#1387106)
  • Fixed starting and stopping cluster services on systemd machines without the service executable (ghissue#115)

[0.9.154] - 2016-09-21

  • There is no change log for this and previous releases. We are sorry.
  • Take a look at git history if you are interested.