Chapter 12. Tungsten Dashboard User Interface

This section describes all of the features and functionality available in our browser-based Graphical User Interface.

12.2. Dashboard Navigation Bar One

Nav Bar One is the first horizontal bar across the top of the window.

Figure 12.2. Example Navigation Bar One

Dashboard Navigation Bar One

  1. Logo and site title - click either to return to the home page (full page load)

  2. Clusters menu - All cluster configured in the config.php file will be displayed in a hierachical view. Click on any onw to limit the view to that cluster. If you select a Composite cluster, the parent and all member clusters will show.

  3. Tools menu - various links to outside resources. Custom links may be added here via the config.php file in the web root directory.

  4. Help feature - click to reveal helpful information.

12.3. Dashboard Navigation Bar Two

Nav Bar Two is the second horizontal bar across the top of the window.

Figure 12.3. Example Navigation Bar Two

Dashboard Navigation Bar Two

The badges for "Policy Not Auto" and "Not Ready" tabs are auto-updated via AJAX every 30 seconds independently of the Auto-Refresh setting on Navigation Bar Two.

  1. All Clusters Tab - click to see all available clusters, same as clicking logo and site title (full page load)

  2. Policy Not Auto Tab - click to see all only those clusters where the policy is set to other than Automatic

  3. Not Ready Tab - click to see only clusters that are not in the Ready state

  4. Filtering feature - enter a value to search for in the cluster name. The search is case in-sensitive and has automatic wildcards on both sides of the string. Click on the Clear button to empty out the filter field.

12.4. Dashboard Navigation Bar Three

Nav Bar Three is the third horizontal bar across the top of the window.

Figure 12.4. Example Navigation Bar Three

Dashboard Navigation Bar Three

  1. Content title - shows current view or filter in use

  2. Auto-refresh feature - select a refresh rate of 0 (off), 5, 10, 30, 60, 120 or 300 seconds. This will enable AJAX-based reloads of the clusters in the content section without reloading the entire page. Look for the spinner in the refresh button per cluster when the refresh is triggered.

  3. Reload button - same as clicking the top logo (full page load)

  4. Hide All Details button - each database node is expandable to display all available details. This button closes them all.

  5. Show All Details button - each database node is expandable to display all available details. This button opens them all.

  6. Collapse All button - each Composite cluster is expandable to display all available node rows. This button closes them all.

  7. Expand All button - each Composite cluster is expandable to display all available node rows. This button opens them all.

  8. Clear Messages button - dismiss all messages that are showing at the top of the screen.

12.5. Dashboard Composite Parent Row

A composite Parent row contains controls for the entire Composite cluster.

Figure 12.5. Example Composite Parent Row

Dashboard Composite Parent Row

  1. Cluster type composite vertical tag, resource icon and parent cluster name

  2. Composite cluster status. The color will change based on the status. Status will be one of: Ready, Warning, or Error

  3. Cluster Policy. One of: Automatic, Maintenance or Mixed. There is a state-sensitive dropdown menu to allow the Policy to be changed.

  4. Cluster type - one of: Standalone, Composite, Master or Slave. Master and Slave both imply Composite membership. Standalone has no composite membership. This field is a duplicate of the vertical tag at the start of field (1), above.

  5. Connections - display the total number of active connections from all Connectors to all nodes in the entire Composite cluster. If you hiver over the info icon, you can see the full breakdown by node.

  6. Composite actions dropdown menu - these are the same commands available when using cctrl -multi followed by use {composite_service_name_here}, i.e.:

    shell> cctrl -multi
    [LOGICAL] / > use world
    [LOGICAL] /world > {your_selected_command_here}
    • Heartbeat (actually cluster heartbeat (in [Tungsten Clustering for MySQL 5.0 Manual]))

    • Recover

    • Switch

    • Failover

  7. Locking status text and icon with dropdown menu to allow lock control.

    Important

    Under normal circumstances, you should not need to get a lock, since all operations automatically attempt to obtain a lock for efficiency purposes.

  8. Refresh button - triggers an AJAX refresh of the parent cluster and all member clusters including all node rows. (no page load)

  9. Collapse all in Composite cluster - hide node rows for all member clusters in this Composite.

12.6. Dashboard Composite Member Rows

A composite member row contains controls for all nodes in the member cluster.

Member clusters may have either the Master or Slave role.

There will be only one Master member cluster and any number of slave member clusters.

Figure 12.6. Example Composite Member Rows

Dashboard Composite Member Rows

  1. Cluster type vertical tag (master or slave), member cluster right-arrow indicator and cluster resource icon

  2. Cluster parent service name followed by the cluster service name

  3. Composite member cluster status. The color will change based on the status. Status will be one of: Ready, Warning, or Error

  4. Cluster Policy. One of: Automatic, Maintenance or Mixed. There is a state-sensitive dropdown menu to allow the Policy to be changed.

  5. Cluster type - one of: Standalone, Composite, Master or Slave. Master and Slave both imply Composite membership. Standalone has no composite membership. This field is a duplicate of the vertical tag at the start of field (1), above.

  6. Connections - display the total number of active connections from all Connectors to all nodes in the entire Composite cluster. If you hiver over the info icon, you can see the full breakdown by node.

  7. Coordinator - display the host which currently has the coordinator role for the member cluster. Every cluster designates one of the Tungsten Managers in the cluster as the coordinator and it is this Manager that will be responsible for taking action, if action is required, to recover the cluster's database resources to the most highly available state possible.

  8. Cluster actions dropdown menu - there are three distinct types of choices in this dropdown menu

    • UI-Specific

      • Toggle Details - show or hide the node details for all nodes in the member cluster

    • Cluster-level commands

      These are the same commands available when using cctrl (in [Tungsten Clustering for MySQL 5.0 Manual]), i.e.:

      shell> cctrl
      [LOGICAL] /east > {your_selected_command_here}

      Note

      The cluster service name displayed will be the service name of the node you are logged into.

      • Heartbeat

      • Recover

      • Switch

      • Failover

    • Composite datasource-level commands

      These are the same commands available when using cctrl -multi followed by use {composite_service_name_here}, i.e.:

      shell> cctrl -multi
      [LOGICAL] / > use world
      [LOGICAL] /world > datasource {cluster_member_service_here} {your_selected_command_here}

      Here are some individual examples:

      [LOGICAL] /world > datasource east recover
      [LOGICAL] /world > datasource west fail
      [LOGICAL] /world > switch to west
      • Recover

      • Welcome

      • Online

      • Offline

      • Shun

      • Promote - this is the same as doing a switch to {cluster_member_service_here}

      • Fail

  9. Locking status text and icon with dropdown menu to allow lock control.

    Important

    Under normal circumstances, you should not need to get a lock, since all operations automatically attempt to obtain a lock for efficiency purposes.

  10. Refresh - triggers an AJAX refresh of that member cluster only (no page load)

  11. Collapse - hide the node rows for that member cluster only

12.7. Dashboard Composite Member Node Rows

A node row contains controls for that one specific cluster node.

Cluster nodes may have one of the following roles: Master, Slave, Witness or Standby. Composite member cluster nodes may also have the Relay role.

For any cluster, there will be only one Master/Relay cluster node and any number of Slave nodes.

A Cluster Master node is assigned the special role of Relay when it is part of a Composite Slave cluster.

Active witness nodes do not have a database and therefore do not run a replicator. Passive witness nodes do not appear because they have no Manager process running.

Figure 12.7. Example Composite Member Node Rows

Dashboard Composite Member Node Rows

  • Node - the hostname of the server

  • Role - one of Master, Relay, Slave, Standby or Witness

  • DS State - DataSource state can be ONLINE (in [Tungsten Clustering for MySQL 5.0 Manual]), OFFLINE (in [Tungsten Clustering for MySQL 5.0 Manual]), SHUNNED or FAILED. There may be other, less-used values.

  • Conns - number of active connections / total number of connections created since last restart

  • Archive - has Archive mode been enabled? See Mark a Datasource as Archive for more information.

  • Repl. State - the state of the Replicator process, one of: ONLINE (in [Tungsten Clustering for MySQL 5.0 Manual]), OFFLINE (in [Tungsten Clustering for MySQL 5.0 Manual]) or ERROR

  • applied - the appliedLatency (in [Tungsten Clustering for MySQL 5.0 Manual]) value, which is how long it took to actually get the event either extracted from the Master's binary logs or applied into the Slave's target database

  • relative - the relativeLatency (in [Tungsten Clustering for MySQL 5.0 Manual]) value, which is how long it has been since we performed an action

  • Seqno - the appliedLastSeqno (in [Tungsten Clustering for MySQL 5.0 Manual]) value

  • minStored - the minimumStoredSeqNo (in [Tungsten Clustering for MySQL 5.0 Manual]) value, which is the sequence number of the oldest event stored in the THL

  • maxStored - the maximumStoredSeqNo (in [Tungsten Clustering for MySQL 5.0 Manual]) value, which is the sequence number of the latest event to be stored in the THL

  • pipelineSource (in [Tungsten Clustering for MySQL 5.0 Manual]) - the protocol, host and port where the replicator is pulling THL from

  • Dataserver - the state of the database server, one of ONLINE (in [Tungsten Clustering for MySQL 5.0 Manual]), OFFLINE (in [Tungsten Clustering for MySQL 5.0 Manual]) or UNKNOWN

  • Actions - the node-specifc commands dropdown menu. There are four distinct types of choices in this dropdown menu.

    • UI-Specific

      • For all nodes that have a running Replicator, the installed Tungsten version will be the first item visible.

      • Toggle Details - show or hide the node details for that specific node

    • DataSource (Node-level) Commands

      These are the same commands available when using cctrl (in [Tungsten Clustering for MySQL 5.0 Manual]), i.e.:

      shell> cctrl
      [LOGICAL] /east > datasource {node_hostname_here} {your_selected_command_here}

      Note

      The cluster service name displayed will be the service name of the node you are logged into.

      • Recover

      • Welcome

      • Offline - only appears if the DataSource is in the ONLINE (in [Tungsten Clustering for MySQL 5.0 Manual]) state

      • Online - only appears if the DataSource is in the OFFLINE (in [Tungsten Clustering for MySQL 5.0 Manual]) state

      • Fail

    • Replicator-specific DataSource (Node-level) Commands

      These are the same commands available when using cctrl (in [Tungsten Clustering for MySQL 5.0 Manual]), i.e.:

      shell> cctrl
      [LOGICAL] /east > replicator {node_hostname_here} {your_selected_command_here}

      Here are some individual examples:

      [LOGICAL] /world > replicator db1 online
      [LOGICAL] /world > replicator db3 offline
      • Offline - only appears if the Replicator is in the ONLINE (in [Tungsten Clustering for MySQL 5.0 Manual]) state

      • Online - only appears if the Replicator is in the OFFLINE (in [Tungsten Clustering for MySQL 5.0 Manual]) state

    • Slave-specific DataSource (Node-level) Commands

      Important

      These are commands are ONLY available on a node with the Slave or Standby roles. Nodes with either Master, Relay or Witness roles will not display the Slave-specific menu options.

      These are the same commands available when using cctrl (in [Tungsten Clustering for MySQL 5.0 Manual]), i.e.:

      shell> cctrl
      [LOGICAL] /east > datasource {node_hostname_here} {your_selected_command_here}

      Here are some individual examples:

      [LOGICAL] /world > datasource db1 shun
      [LOGICAL] /world > datasource db3 recover
      [LOGICAL] /world > switch to db2
      • Backup

      • Restore

      • Shun

      • Enable Standby

      • Disable Standby

      • Promote - this is the same as doing a switch to {node_hostname_here}

12.8. Dashboard Standalone Cluster

All of the controls and information are the same for Standalone clusters and nodes as they are for Composite with the following exceptions:

  • A Standalone Cluster is not part of a Composite.

  • There will be no Composite commands in the service-level dropdown menu.

Figure 12.8. Example Standalone Cluster

Dashboard Standalone Cluster