4.1 KiB

Raw Blame History

Active Failover Deployment

This Section describes how to deploy an Active Failover environment. For a general introduction to Active Failover, please refer to the Active Failover chapter.

As usual there are two main ways to start an Active-Failover setup: Either manually or using the ArangoDB Starter (possibly in conjunction with docker).

Starting Manually

We are going to start two single server instances and one Agency. First we need to start the

    arangod \
        --agency.activate true \
        --agency.endpoint tcp://agency.domain.org:4001 \
        --agency.my-address tcp://agency.domain.org:4001 \
        --agency.pool-size 1 \
        --agency.size 1 \
        --database.directory dbdir/data4001 \
        --javascript.v8-contexts 1 \
        --server.endpoint tcp://agency.domain.org:4001 \
        --server.statistics false \
        --server.threads 16 \
        --log.file dbdir/4001.log \
        --log.level INFO \
        | tee dbdir/4001.stdout 2>&1 &

Next we are going to start the leader (wait until this server is fully started)

    arangod \
      --database.directory dbdir/data8530 \
      --cluster.agency-endpoint tcp://agency.domain.org:4001 \
      --cluster.my-address tcp://leader.domain.org:4001 \
      --server.endpoint tcp://leader.domain.org:4001 \
      --cluster.my-role SINGLE \
      --replication.active-failover true \
      --log.file dbdir/8530.log \
      --server.statistics true \
      --server.threads 5 \
      | tee cluster/8530.stdout 2>&1 &

After the leader server is fully started then you can add additional followers, with basically the same startup parameters (except for their address and database directory)

    arangod \
      --database.directory dbdir/data8531 \
      --cluster.agency-endpoint tcp://agency.domain.org:4001 \
      --cluster.my-address tcp://leader.domain.org:4001 \
      --server.endpoint tcp://leader.domain.org:4001 \
      --cluster.my-role SINGLE \
      --replication.active-failover true \
      --log.file dbdir/8531.log \
      --server.statistics true \
      --server.threads 5 \
      | tee cluster/8531.stdout 2>&1 &

Using the ArangoDB Starter

If you want to start a resilient single database server, use --starter.mode=resilientsingle. In this mode a 3 machine _Agency is started as well as 2 single servers that perform asynchronous replication an failover, if needed.

arangodb --starter.mode=resilientsingle --starter.join A,B,C

Run this on machine A, B & C.

The Starter will decide on which 2 machines to run a single server instance. To override this decision (only valid while bootstrapping), add a --cluster.start-single=false to the machine where the single server instance should NOT be scheduled.

Starting a resilient single server pair in Docker

If you want to start a resilient single database server running in docker containers, use the normal docker arguments, combined with --starter.mode=resilientsingle.

export IP=<IP of docker host>
docker volume create arangodb
docker run -it --name=adb --rm -p 8528:8528 \
    -v arangodb:/data \
    -v /var/run/docker.sock:/var/run/docker.sock \
    arangodb/arangodb-starter \
    --starter.address=$IP \
    --starter.mode=resilientsingle \
    --starter.join=A,B,C

Run this on machine A, B & C.

The starter will decide on which 2 machines to run a single server instance. To override this decision (only valid while bootstrapping), add a --cluster.start-single=false to the machine where the single server instance should NOT be scheduled.

Starting a local test resilient single sever pair

If you want to start a local resilient server pair quickly, use the --starter.local flag. It will start all servers within the context of a single starter process.

arangodb --starter.local --starter.mode=resilientsingle

Note: When you restart the started, it remembers the original --starter.local flag.

4.1 KiB Raw Blame History