Using Failover Manager

|variable_prod_name| offers support for monitoring and failover of clusters with one or more Standby servers. You can add or remove nodes from the cluster as your demand for resources grows or shrinks.

If a Master node reboots, |variable_prod_name| may detect the database is down on the Master node and promote a Standby node to the role of Master. If this happens, the |variable_prod_name| agent on the (rebooted) Master node will not get a chance to write the recovery.conf file (for server version 11 or prior) or standby.signal file (for server version 12 or later); the rebooted Master node will return to the cluster as a second Master node. To prevent this, start the |variable_prod_name| agent before starting the database server. The agent will start in idle mode, and check to see if there is already a master in the cluster. If there is a master node, the agent will verify that a recovery.conf or standby.signal file exists, and the database will not start as a second master.

Managing a Failover Manager Cluster

Once configured, a |variable_prod_name| cluster requires no regular maintenance. The following sections provide information about performing the management tasks that may occasionally be required by a Failover Manager Cluster.

By default, some of the efm commands must be invoked by efm or an OS superuser; an administrator can selectively permit users to invoke these commands by adding the user to the efm group. The commands are:

Starting the Failover Manager Cluster

You can start the nodes of a |variable_prod_name| cluster in any order.

To start the |variable_prod_name| cluster on RHEL 6.x or CentOS 6.x, assume superuser privileges, and invoke the command:

service edb-efm-|variable_prod_version| start

To start the |variable_prod_name| cluster on RHEL/CentOS 7.x or RHEL/CentOS 8.x, assume superuser privileges, and invoke the command:

systemctl start edb-efm-|variable_prod_version|

If the cluster properties file for the node specifies that is.witness is true, the node will start as a Witness node.

If the node is not a dedicated Witness node, |variable_prod_name| will connect to the local database and invoke the pg_is_in_recovery() function. If the server responds false, the agent assumes the node is a Master node, and assigns a virtual IP address to the node (if applicable). If the server responds true, the |variable_prod_name| agent assumes that the node is a Standby server. If the server does not respond, the agent will start in an idle state.

After joining the cluster, the |variable_prod_name| agent checks the supplied database credentials to ensure that it can connect to all of the databases within the cluster. If the agent cannot connect, the agent will shut down.

If a new master or standby node joins a cluster, all of the existing nodes will also confirm that they can connect to the database on the new node.

Adding Nodes to a Cluster

You can add a node to a |variable_prod_name| cluster at any time. When you add a node to a cluster, you must modify the cluster to allow the new node, and then tell the new node how to find the cluster. The following steps detail adding a node to a cluster:

  1. Unless auto.allow.hosts is set to true, use the efm allow-node command, to add the IP address of the new node to the Failover Manager allowed node host list. When invoking the command, specify the cluster name and the IP address of the new node:

    efm allow-node <cluster_name ip_address>

    For more information about using the efm allow-node command or controlling a |variable_prod_name| service, see Using the EFM Utility.

    Install a |variable_prod_name| agent and configure the cluster properties file on the new node. For more information about modifying the properties file, see The Cluster Properties File.

  2. Configure the cluster members file on the new node, adding an entry for the Membership Coordinator. For more information about modifying the cluster members file, see The Cluster Members File.

  3. Assume superuser privileges on the new node, and start the Failover Manager agent. To start the |variable_prod_name| cluster on RHEL 6.x or CentOS 6.x, assume superuser privileges, and invoke the command:

service edb-efm-|variable_prod_version| start

To start the |variable_prod_name| cluster on RHEL/CentOS 7.x or RHEL/CentOS 8.x, assume superuser privileges, and invoke the command:

systemctl start edb-efm-|variable_prod_version|

When the new node joins the cluster, |variable_prod_name| will send a notification to the administrator email provided in the user.email property, and/or will invoke the specified notification script.

Please note: : To be a useful Standby for the current node, the node must be a standby in the PostgreSQL Streaming Replication scenario.

Changing the Priority of a Standby

If your |variable_prod_name| cluster includes more than one Standby server, you can use the efm set-priority command to influence the promotion priority of a Standby node. Invoke the command on any existing member of the |variable_prod_name| cluster, and specify a priority value after the IP address of the member.

For example, the following command instructs |variable_prod_name| that the acctg cluster member that is monitoring 10.0.1.9 is the primary Standby (1):

efm set-priority acctg 10.0.1.9 1

You can set the priority of a standby to 0 to make the standby non-promotable. Setting the priority of a standby to a value greater than 0 overrides a property value of promotable=false.

For example, if the properties file on node 10.0.1.10 includes a setting of promotable=false and you use efm set-priority to set the promotion priority of 10.0.1.10 to be the standby used in the event of a failover, the value designated by the efm set-priority command will override the value in the property file:

efm set-priority acctg 10.0.1.10 1

In the event of a failover, |variable_prod_name| will first retrieve information from Postgres streaming replication to confirm which Standby node has the most recent data, and promote the node with the least chance of data loss. If two Standby nodes contain equally up-to-date data, the node with a higher user-specified priority value will be promoted to Master unless use.replay.tiebreaker is set to false . To check the priority value of your Standby nodes, use the command:

efm cluster-status <cluster_name>

Please note: : The promotion priority may change if a node becomes isolated from the cluster, and later re-joins the cluster.

Promoting a Failover Manager Node

You can invoke efm promote on any node of a |variable_prod_name| cluster to start a manual promotion of a Standby database to Master database.

Manual promotion should only be performed during a maintenance window for your database cluster. If you do not have an up-to-date Standby database available, you will be prompted before continuing. To start a manual promotion, assume the identity of efm or the OS superuser, and invoke the command:

efm promote <cluster_name> [-switchover] [-sourcenode <address>] [-quiet] [-noscripts]

Where:

<cluster_name> is the name of the |variable_prod_name| cluster.

Include the –switchover option to reconfigure the original Master as a Standby. If you include the –switchover keyword, the cluster must include a master node and at least one standby, and the nodes must be in sync.

Include the –sourcenode keyword to specify the node from which the recovery settings will be copied to the master.

Include the -quiet keyword to suppress notifications during switchover.

Include the -noscripts keyword to prevent instruct |variable_prod_name| to not invoke fencing and post-promotion scripts.

During switchover:

  • For server versions 11 and prior, the recovery.conf file is copied from an existing standby to the master node. For server version 12 and later, the primary_conninfo and restore_command parameters are copied and stored in memory.

  • The master database is stopped.

  • If you are using a VIP, the address is released from the master node.

  • A standby is promoted to replace the master node, and acquires the VIP.

  • The address of the new master node is added to the recovery.conf file or the primary_conninfo details are stored in memory.

  • If the application.name property is set for this node, the application_name property will be added to the recovery.conf file or the primary_conninfo information will be stored in memory.

  • If you are using server version 12 or later, the recovery settings that have been stored in memory are written to the postgresql.auto.conf file.

  • The old master is started; the agent will resume monitoring it as a standby.

During a manual promotion, the Master agent releases the virtual IP address before creating a recovery.conf file in the directory specified by the db.data.dir property. The recovery.conf file is created on all server versions, and is used to prevent the old master database from starting until the file is removed, preventing the node from starting as a second master in the cluster.

The Master agent remains running, and assumes a status of Idle.

The Standby agent confirms that the virtual IP address is no longer in use before pinging a well-known address to ensure that the agent is not isolated from the network. The Standby agent runs the fencing script and promotes the Standby database to Master. The Standby agent then assigns the virtual IP address to the Standby node, and runs the post-promotion script (if applicable).

Please note that this command instructs the service to ignore the value specified in the auto.failover parameter in the cluster properties file.

To return a node to the role of master, place the node first in the promotion list:

efm set-priority <cluster_name> <ip_address> <priority>

Then, perform a manual promotion:

efm promote <cluster_name> ‑switchover

For more information about the efm utility, see Using the EFM Utility.

Stopping a Failover Manager Agent

When you stop an agent, |variable_prod_name| will remove the node’s address from the cluster members list on all of the running nodes of the cluster, but will not remove the address from the |variable_prod_name| Allowed node host list.

To stop the |variable_prod_name| agent on RHEL 6.x or CentOS 6.x, assume superuser privileges, and invoke the command:

service edb-efm-|variable_prod_version| stop

To stop the |variable_prod_name| agent on RHEL/CentOS 7.x or RHEL/CentOS 8.x, assume superuser privileges, and invoke the command:

systemctl stop edb-efm-|variable_prod_version|

Until you invoke the efm disallow-node command (removing the node’s address of the node from the Allowed node host list), you can use the service edb-efm-3.10 start command to restart the node at a later date without first running the efm allow-node command again.

Please note that stopping an agent does not signal the cluster that the agent has failed unless the master.shutdown.as.failure property is set to true.

Stopping a Failover Manager Cluster

To stop a |variable_prod_name| cluster, connect to any node of a Failover Manager cluster, assume the identity of efm or the OS superuser, and invoke the command:

efm stop-cluster <cluster_name>

The command will cause all |variable_prod_name| agents to exit. Terminating the |variable_prod_name| agents completely disables all failover functionality.

Please note: : When you invoke the efm stop-cluster command, all authorized node information is lost from the Allowed node host list.

Removing a Node from a Cluster

The efm disallow-node command removes the IP address of a node from the |variable_prod_name| Allowed Node host list. Assume the identity of efm or the OS superuser on any existing node (that is currently part of the running cluster), and invoke the efm disallow-node command, specifying the cluster name and the IP address of the node:

efm disallow-node <cluster_name> <ip_address>

The efm disallow-node command will not stop a running agent; the service will continue to run on the node until you stop the agent. If the agent or cluster is subsequently stopped, the node will not be allowed to rejoin the cluster, and will be removed from the failover priority list (and will be ineligible for promotion).

After invoking the efm disallow-node command, you must use the efm allow-node command to add the node to the cluster again.

Running Multiple Agents on a Single Node

You can monitor multiple database clusters that reside on the same host by running multiple Master or Standby agents on that |variable_prod_name| node. You may also run multiple Witness agents on a single node. To configure |variable_prod_name| to monitor more than one database cluster, while ensuring that |variable_prod_name| agents from different clusters do not interfere with each other, you must:

  1. Create a cluster properties file for each member of each cluster that defines a unique set of properties and the role of the node within the cluster.

  2. Create a cluster members file for each member of each cluster that lists the members of the cluster.

  3. Customize the service script (on a RHEL or CentOS 6.x system) or the unit file (on a RHEL/CentOS 7.x or RHEL/CentOS 8.x system) for each cluster to specify the names of the cluster properties and the cluster members files.

  4. Start the services for each cluster.

The examples that follow uses two database clusters (acctg and sales) running on the same node:

  • Data for acctg resides in /opt/pgdata1; its server is monitoring port 5444.

  • Data for sales resides in /opt/pgdata2; its server is monitoring port 5445.

To run a |variable_prod_name| agent for both of these database clusters, use the efm.properties.in template to create two properties files. Each cluster properties file must have a unique name. For this example, we create acctg.properties and sales.properties to match the acctg and sales database clusters.

The following parameters must be unique in each cluster properties file:

admin.port

bind.address

db.port

db.data.dir

virtual.ip (if used)

virtual.ip.interface (if used)

Within each cluster properties file, the db.port parameter should specify a unique value for each cluster, while the db.user and db.database parameter may have the same value or a unique value. For example, the acctg.properties file may specify:

db.user=efm_user

db.password.encrypted=7c801b32a05c0c5cb2ad4ffbda5e8f9a

db.port=5444

db.database=acctg_db

While the sales.properties file may specify:

db.user=efm_user

db.password.encrypted=e003fea651a8b4a80fb248a22b36f334

db.port=5445

db.database=sales_db

Some parameters require special attention when setting up more than one |variable_prod_name| cluster agent on the same node. If multiple agents reside on the same node, each port must be unique. Any two ports will work, but it may be easier to keep the information clear if using ports that are not too close to each other.

When creating the cluster properties file for each cluster, the db.data.dir parameters must also specify values that are unique for each respective database cluster.

The following parameters are used when assigning the virtual IP address to a node. If your |variable_prod_name| cluster does not use a virtual IP address, leave these parameters blank.

virtual.ip

virtual.ip.interface

virtual.ip.prefix

This parameter value is determined by the virtual IP addresses being used and may or may not be the same for both acctg.properties and sales.properties.

After creating the acctg.properties and sales.properties files, create a service script or unit file for each cluster that points to the respective property files; this step is platform specific. If you are using RHEL 6.x or CentOS 6.x, see RHEL 6.x or CentOS 6.x; if you are using RHEL/CentOS 7.x or RHEL/CentOS 8.x, see RHEL/CentOS 7.x or RHEL/CentOS 8.x.

Please note: : If you are using a custom service script or unit file, you must manually update the file to reflect the new service name when you upgrade |variable_prod_name|.

RHEL 6.x or CentOS 6.x

If you are using RHEL 6.x or CentOS 6.x, you should copy the edb-efm-3.10 service script to new file with a name that is unique for each cluster. For example:

# cp /etc/init.d/edb-efm-|variable_prod_version| /etc/init.d/efm-acctg

# cp /etc/init.d/edb-efm-|variable_prod_version| /etc/init.d/efm-sales

Then edit the CLUSTER variable, modifying the cluster name from efm to acctg or sales.

After creating the service scripts, run:

# chkconfig efm-acctg on

# chkconfig efm-sales on

Then, use the new service scripts to start the agents. For example, you can start the acctg agent with the command:

# service efm-acctg start

RHEL/CentOS 7.x or RHEL/CentOS 8.x

If you are using RHEL/CentOS 7.x or RHEL/CentOS 8.x, you should copy the edb-efm-3.10 unit file to new file with a name that is unique for each cluster. For example, if you have two clusters (named acctg and sales), the unit file names might be:

/usr/lib/systemd/system/efm-acctg.service

/usr/lib/systemd/system/efm-sales.service

Then, edit the CLUSTER variable within each unit file, changing the specified cluster name from efm to the new cluster name. For example, for a cluster named acctg, the value would specify:

Environment=CLUSTER=acctg

You must also update the value of the PIDfile parameter to specify the new cluster name. For example:

PIDFile=/var/run/efm-|variable_prod_version|/acctg.pid

After copying the service scripts, use the following commands to enable the services:

# systemctl enable efm-acctg.service

# systemctl enable efm-sales.service

Then, use the new service scripts to start the agents. For example, you can start the acctg agent with the command:

# systemctl start efm-acctg

For information about customizing a unit file, please visit:

http://fedoraproject.org/wiki/Systemd#How_do_I_customize_a_unit_file.2F_add_a_custom_unit_file.3F