Changes in v1.4.1

Patch Release: June 20^th, 2017

• Build:
  • clustercontrol-1.4.1-3393
• UI:
  • Fix for a build issue on Ubuntu/Debian.

Patch Release: June 19^th, 2017

• Build:
  • clustercontrol-1.4.1-3384
• UI:
  • Fix for setting the Settings→Backup’s retention period. In future versions Settings, →Backups will be deprecated/removed and can be accessed from the Backup page instead.
  • Fix inconsistent backup executed and next execution time and timezones displayed. UTC timezone is used across the backup page for now.
  • Performance →Transaction Log is disabled as default. Added a slider to set the sampling interval.
  • Add Node and Add Existing Node now has a data directory input field to change the data directory used for the new node.

Patch Release: May 24^th, 2017

• Build:
  • clustercontrol-1.4.1-3181
• UI:
  • The alarm category in the Activity Viewer is now correctly showing the component name instead of the type name.
  • Fix to show the correct server name in the individual server load graphs.
  • Fix regression/empty table for Performance →DB Variables.
  • Fix to enable editable dropdown to the Add Existing Keepalived form for HAProxy.
  • Support for using a custom port when adding a MySQL Asynchronous Slave (MySQL Replication)
  • Fix for Configuration Management →Change to list only valid nodes.
  • Performance →Status Time Machine is now deprecated/removed.

 Patch Release: May 20^th, 2017

• Build:
  • clustercontrol-controller-1902
• Controller:
  • Disable by default tx deadlock detection as it takes a lot of CPU. Added new param: db_deadlock_check_interval
  • How often to check for deadlocks. 0 means disabled (default). Specified in seconds. (default: 0).
  • Enable in /etc/cmon.d/cmon_X.cnf (if you want to enable it, then 20 is a good value) and restart cmon.
  • Sample controller IP is seen by MySQL nodes once after every cmon restart.
    logrotate (wtmp) more often and restart accounts-daemon
  • A fix of show_db_users and show_db_unusued_accounts java scripts.

Patch Release: May 12^th, 2017

• Build:
  • clustercontrol-1.4.1-3121
  • clustercontrol-controller-1.4.1-1890
  • clustercontrol-cmonapi-274
• UI:
  • ProxySQL: Fix the wrong IP in proxysql selected node header.
  • PostgreSQL fixes:
    • Overview page no longer causes high load on the web client.
    • Performance → DB Variables are now loading up correctly.
    • Tooltips added for the graphs.
  • LDAP authentication attempts are logged to a separate log file, {webdir}/clustercontrol/app/log/cc-ldap.log
  • Minor improvements on how multiple recipients for email notifications are added.
• Controller:
  • Galera: Fixed a bug in the clone cluster
  • Deployment: Fixed a bug using hostnames, which could cause grant/privilege errors from the controller preventing the controller to connect to the managed nodes.
  • ProxySQL: hashing of passwords in the mysql_users table.
  • Backup Reports: Properly transform IP’s into hostnames in the backup report (due to a previous UI bug, some backups&schedules are used IPs instead of hostnames).
  • MongoDB: Degraded cluster state reported after removing shard.
• CMONAPI:
  • Fixed an issue causing not all recipients to be listed under Settings (top menu) → Email Notifications.

 Patch Release: Apr 24^th, 2017

• Build:
  • clustercontrol-1.4.1-304
  • clustercontrol-controller-1.4.1-1856
  • clustercontrol-cmonapi-266
• UI:
  • Fix for empty databases list with MySQL backups.
  • MySQL Variables page now uses the RPC API.
  • Improved deployment wizard placeholders descriptions.
  • Enable ‘restore backup’ for PostgreSQL.
  • Enable using a custom PostgreSQL port (default 5432) for deployments.
  • Fix for allowing negative port numbers in the load balancer forms.
  • Fix empty details on the keepalived node page.
  • Fix for saving timezone settings other than GMT+0 with email notifications.
• Controller:
  • Fix for deploying a single MySQL replication node cluster.
  • Require set force to stop a read-write MySQL server (MySQL Replication).
  • Fix for node(s) reconnection issue to restored master after a restore backup.
  • Fix configuration (my.cnf) import to start immediately after a MySQL replication slave has been added (Galera)
  • Job log improvement. Show the command/action that was requested.
  • Fix with MaxScale to show the correct list of masters and slaves in the console.

Patch Release: Apr 12th, 2017

• Build:
  • clustercontrol-1.4.1-3002
  • clustercontrol-controller-1.4.1-1834
• UI:
  • Fixed a bug making it impossible to restart failed jobs.
  • Fixed a bug in the Nodes graphs which made them render wrongly
  • Replication: Extended the Import dialog (Replication cluster) with a few more options (enable information schema queries).
  • Galera: Added Multi Nic support for Add Replication Slave.
    Fixed the title for the Nodes page.
  • ProxySQL: Handle latency (us/ms) and improvements to graphs.
  • Query Monitor: Top queries useless with more than 20 queries.
  • New Operation Report – Schema Change Report. With this feature, you can spot changes in your database schemas and ensure changes are sound on your system.
• Controller:
  • Fixed a bug making it impossible to add an existing replication slave.
  • Replication (Percona,MySQL): print out messages to show progress while applying relay log.
  • Java script fixes to take the enable_is_queries setting into account
  • SSH alarms re-organized and an alarm is raised if SSH access is determined to be too slow.
  • GroupRepl: fixing add-replication slave bug
  • A JS script to change the password on all MySQL servers (mainly useful only for NDB)
  • ProxySQL: small fix for ‘latency’. Older versions used Latency_ms, newer Latency_us.
  • User option: enable_is_queries = 0|1
  • Detect schema changes (CREATE and ALTER TABLE. Drop table is not supported yet). New options: schema_change_detection_address, schema_change_detection_databases, schema_change_detection_pause_time_ms must be set in /etc/cmon.d/cmon_X.cnf to enable the feature. A new Operation Report (Schema Change) must be scheduled.
  • Creating a report of 100 000 schemas and tables will take about 5-10 minutes depending on hardware. Configure the schema_change_detection_address to run on a replication slave or an async slave connected to e.g a Galera or Group Replication Cluster. For NDB this schema_change_detection_address should be set to a MySQL server used for admin purposes. Throttle the detection process with schema_change_detection_pause_time_ms. schema_change_detection_databases is a comma-separated string of database names and also supports wildcards, e.g ‘DB%’, will evaluate all database starting with DB.

Initial Release: Apr 4^th, 2017

• Build:
  • clustercontrol-1.4.1-2967
  • clustercontrol-cmonapi-1.4.1-257
  • clustercontrol-nodejs-1.4.1-86
  • clustercontrol-controller-1.4.1-1811

In this release, we have added additional management functions for ProxySQL. You can now view queries passing through ProxySQL, create and edit query rules, host groups/servers, users, and variables. We also have support for managing MySQL Galera and Replication clusters using separate management and data/database IPs for improved security.

• ProxySQL (v2):
  • Support for MySQL Galera in addition to Replication clusters.
  • Support for active-standby HA setup with KeepAlived.
  • Use the Query Monitor to view query digests.
  • Manage Query Rules (Query Caching, Query Rewrite).
  • Manage Host Groups (Servers).
  • Manage ProxySQL DB Users.
  • Manage ProxySQL System Variables.
• Multiple Networks/Segmented Traffic
  • Manage MySQL Galera and Replication clusters with management/public IPs for monitoring connections and data/private IPs for replication traffic.
  • Add Galera nodes or Replication Read Slaves with management and data IPs.