2. Home
  4. Docs
  6. ClusterControl
  8. Changelogs
  10. Changes in v1.5.1

Changes in v1.5.1

Maintenance Release: April 9th, 2018

  • Build:
    • clustercontrol-controller-1.5.1-2467
  • Controller:
    • Monitoring: SSH Optimizations to reduce the number of SSH connections on remote nodes.
    • Monitoring: CPU temperature monitoring is now configurable (and disabled by default, monitor_cpu_temperature cmon configuration option).
    • Galera: Disable P_S queries in Query Monitor during the upgrade.
    • Galera: Add node – check if MariaDB version is of 10.1.31 and above. In this case, mariabackup will be used.
    • ProxySQL: Fixed an issue when modifying the variable values from the UI.
    • MaxScale: Template issue with a configuration parameter not compatible with MySQL Monitor module.
    • Maxscale: Debian 9 support.
    • HAProxy: If xinetd failed to install it could lead to the controller crashing.
    • Fixing a license barrier when deploying Galera cluster causing an error: Refusing to recover node (no license)
    • Mariadb 10.1 now requires wsrep_sst_method=mariabackup (new deploys of MariaDB will always use mariabackup for SST).

Maintenance Release: March 7th, 2018

  • Build:
    • clustercontrol-controller-1.5.1-2411
    • clustercontrol-1.5.1-4434
  • Frontend/UI:
    • CRITICAL: Fixed another issue where the wrong node was selected due to an indexing problem, which could lead to an action being executed on the wrong node.
  • Controller:
    • Fixed an issue when importing keepalived.

Maintenance Release: March 6th, 2018

  • Build:
    • clustercontrol-controller-1.5.1-2409
    • clustercontrol-1.5.1-4425
  • Controller:
    • PostgreSQL: Explicitly grant nodes by IP (in addition to hostnames) in pg_hba.conf.
    • PostgreSQL: config writes with includes caused invalid syntax error issues.
    • MySQL Cluster: Bug fixes to Database Growth.
    • Operational Reports: Improved handling of different gnuplot versions.
    • General: Configurable ICMP pinging. Set enable_icmp_ping=false to disable ICMP pinging (Azure requires this). By default it is true (recommended).
  • Frontend/UI:
    • Installer: Permissions fixed so there are no writable files after install.
    • Fixed an issue where the wrong node was selected due to an indexing problem, which could lead to an action being executed on the wrong node.
    • Improved handling of saving email notification settings.

Maintenance Release: Feb 24th, 2018

  • Build:
    • clustercontrol-controller-1.5.1-2390
    • clustercontrol-1.5.1-4395
  • Controller:
    • PostgreSQL: Fixed a bug in the PostgreSQL config parsing causing a syntax error using include.
    • Advisors: Bug fixes and corrections
    • MySQL Cluster: Fixed a number of issues around hostnames and port settings, which caused node types (data node, management node) to be improperly identified.
    • Backup (Verify Backup): Fixed a number of issues handling the Backup Verification Server.
    • Backup (Verify Backup): A backup verification email is now sent when the backup has been verified.
    • Operational Reports: Availability report issues. The cluster_events/node_events tables were inadvertently dropped during ClusterControl upgrades causing the stats to be reset.
    • PostgreSQL: pg_basebackup executed on a slave failed on imported clusters due to a missing grant.
    • Remove Node: Fixes to make it possible to only unregister a node (remove the node from ClusterControl management).
    • Schema: DB schema fixes to the server_node properties column by extending the size.
    • Galera/Group Replication: properly write/update the cmon_X.cnf mysql_server_addresses field to mark non galera|group_repl nodes there correctly.
  • Frontend/UI:
    • Remove Node: Improved consistency in ‘Remove Node’’ dialogs.
    • MySQL/Galera: New default value for binary logging path which is now outside of the datadir.
    • Backup (MySQL based clusters): Fixed an issue where Backup Method and Backup Host drop-downs were empty.
    • Deploy/Import/Add Node: Improved Host Discovery showing the actual SSH error.
    • Deploy/Import: SSH Key Path validation was missing.
    • Charts: It was possible to select a negative range (smaller end date than the start date).
    • MongoDB: Add Shards dialog got stuck when entering a hostname (SSH check never terminated).

 Maintenance Release: Feb 6th, 2018

  • Build:
    • clustercontrol-controller-1.5.1-2362
    • clustercontrol-1.5.1-4356
  • Controller:
    • Stats: RAM stats updated to also account for SReclaimable.
    • PostgreSQL: enable pg_stats_statements extension only on non-read-only nodes.
    • Error Reporter: include more info about PostgreSQL clusters (pg_stat_replication table + recovery config file)
    • MySQL: Fixed an issue handling !include directives containing quotes. The import config job (automatically executed upon a controller restart) will auto-correct broken MySQL !include directives containing quotes.
  • Frontend/UI:
    • Deployment/Import dialogs: Added validation for SSH Key Path.
    • ProxySQL: Filter users with unrelated hosts when deploying ProxySQL.
    • MongoDB: Fixed a problem specifying hostnames when performing Add Shard.

Maintenance Release: Jan 29th, 2018

  • Build:
    • clustercontrol-controller-1.5.1-2346
  • Controller
    • Galera: A path to a node_recovery_lock_file can now be specified in /etc/cmon.d/cmon_N.cnf. If set and the lock file is found on the node then the node recovery will fail until an admin/script removes this file. The cmon controller process must be restarted when this parameter has been specified. This feature may be useful for encrypted filesystems.
    • MySQL Cluster (NDB): Fix to allow deployments on other ports than 3306.
    • MySQL: Error code 2013 (lost connection during query) is not a reason to set a node into a disconnected state.
    • MySQL: Fixed up handling of ignore-db-dir in config templates on MySQL 5.5 based servers
    • ProxySQL: Improve ProxySQL support such that admin-admin_credentials may contain multiple credentials.

 Maintenance Release: Jan 23rd, 2018

  • Build:
    • clustercontrol-1.5.1-4335
    • clustercontrol-controller-1.5.1-2335
  • Controller:
    • Load balancers: Fix to make it possible to remove HAProxy/MaxScale even if the host is not reachable.
    • Error reporting: Fix to always include cluster-id 0 jobs in the error report.
    • Galera: Fix to disallow garbd deployment to a host having a running mysqld.
    • Replication: Improve handling of read_only when importing existing replication cluster.
    • Replication: Alert if a MySQL server is not connected to any master, i.e hanging loose.
    • Postgres: Fix to recover a failed PostgreSQL server in case there is only one single Postgres node in the system.
    • Postgres: Fix to prevent PostgreSQL to be restarted in case sending SIGHUP (to reload config) failed.
    • Advisors: Fix to present a clear error message for the performance schema advisors in case performance schema
    • tables are not available for a particular MySQL version.
    • Verify Backup: Fix to correctly stage the standalone node with MySQL user info (cmon user, etc).
  • UI:
    • Fix properly enabling read/write split with HAProxy and MySQL Galera.
    • Fix incorrect list of nodes showing up as bootstrapping candidates (Galera).
    • Fix leaving user records behind when deleting the whole team.
    • Add an option to limit network streaming bandwidth (Mb/s) when doing a backup.
    • Fix missing read-only port when adding HAProxy for PostgreSQL.
    • Fix showing the correct read/write port when adding HAProxy for PostgreSQL.
    • Fix Query Monitor for PostgreSQL to show the complete query and not truncate it.
    • Fix misleading tooltips when deploying or importing a PostgreSQL cluster.
    • Remove the requirement to have the binlog enabled when adding a SQL Node with MySQL NDB Cluster.
    • Remove incorrect software package option when adding a SQL Node” with MySQL NDB Cluster.
    • Fix MaxScale console port issue with using Safari.
    • Fix schedule backups to work even when the verify backup option is enabled.

Initial Release: Dec 22nd, 2017

  • Build:
    • clustercontrol-1.5.1-4265
    • clustercontrol-controller-1.5.1-2299
    • clustercontrol-cmonapi-1.5.0-290
    • clustercontrol-notifications-1.5.0-70
    • clustercontrol-ssh-1.5.0-39
    • clustercontrol-cloud-1.5.0-31
    • clustercontrol-clud-1.5.0-31

In this release, we have added support to optionally use our built-in AES-256 encryption for your backups. Secure your backups for offsite or cloud storage with a flip of a checkbox.

We have also added an option to use a custom retention period per the backup schedule.

There is a new Topology view (BETA) initially with MySQL based clusters to show a replication topology (incl. any load balancers) for your cluster. Use drag and drop to perform node actions, for example, drag a replication slave on top of a master node which will prompt you to either rebuild the slave or change the replication master.

A new left-side navigation bar provides faster page access to some of our features and the node actions are now also accessible directly from the node list.

Feature Details

    • AES-256 Backup Encryption (and Restore):
      • Supported backup methods:
        • mysqldump, xtrabackup (MySQL).
        • pg_dump, pg_basebackup (PostgreSQL).
        • mongodump (MongoDB).
    • Topology View (BETA):
      • MySQL Replication Topology.
      • MySQL Galera Topology.
    • MongoDB
      • Support for MongoDB v3.4.
      • Fix to add back restore from backup.
      • Multiple NICs support. Management/public IPs for monitoring connections and data/private IPs for replication traffic.
    • Misc:
      • Left side navigation.
      • Global settings breakout.
      • Quick node actions.
Was this article helpful to you? Yes No