Skip to main content

loadbalancerstatus

Use loadbalancerstatus to check node health when using a load balancer.

A scheduled job runs loadbalancerstatus with a -clear option to trim the contents of the records table of a db in db/nodestatus to remove all entries older than some number of days in the past from the time of running.

This can be configured through LOADBALANCERSTATUS RECORD HIST DAYS in Manage the system > Maintenance > Options. The default, when this system variable is not set, is 30 days.

Usage
loadbalancerstatus <options>

Argument

Description

-disk

Check remaining disk space. Returns success if the node’s replication queues have not hit the high water mark.

-db_commit_suspend

Check if the instance is in DB commit suspend.

Returns success unless this node has suspended database commit operations, which typically occur during resynchronization or when a replication queue has become full.

-services <services>

Check if services are running. Note that iddb and ajaxsvc must be running for any node status test to succeed.

-ping

Check if iddb can connect to its backend database.

-plugin <plugin>

Run this plugin from the plugins directory.

-record <filename>

Record statistics to this database in the db/nodestatus directory.

-down

Always fail (use this to manually keep a node out of a load balancer).

-ajax_max_timeouts <N>

Maximum number of Ajax timeouts to allow.

-ajax_timeout_minutes <N>

Number of minutes to observe Ajax timeouts.

-clear

Clear records for specified filename to last <x> days.

The number can be configured through LOADBALANCERSTATUS RECORD HIST DAYS in Manage the system > Maintenance > Options . The default, when this system variable is not set, is 30 days.

-clear_all

Clear records for all files to last <x> days.

The number can be configured through LOADBALANCERSTATUS RECORD HIST DAYS in Manage the system > Maintenance > Options . The default, when this system variable is not set, is 30 days.

In this section: