ArangoDB Server Agency Options

Activate

agency.activate

Activate agency.

Compaction

agency.compaction-keep-size

Keep as many indices before compaction point.

Election

agency.election-timeout-max

Maximum timeout before an agent calls for new election in seconds.

agency.election-timeout-min

Minimum timeout before an agent calls for new election in seconds.

Endpoint

agency.endpoint

Agency endpoints.

My address

agency.my-address

Which address to advertise to the outside.

Pool size

agency.pool-size

Number of agent pool.

Size

agency.size

Number of agents.

Supervision

agency.supervision

Perform ArangoDB cluster supervision.

agency.supervision-frequency

ArangoDB cluster supervision frequency in seconds.

agency.supervision-grace-period

Supervision time, after which a server is considered to have failed, in seconds.

The default value is 10 seconds, which is recommended for regular cluster deployments. For active failover deployments it is recommended to use a higher value for the grace period to avoid unnecessary failovers. In active failover setups, the leader server needs to handle all the load and is thus expected to get overloaded and unresponsive more easily than a server in a regular cluster which needs to handle only a part of the overall load.