With the upcoming end-of-life (EOL) of MSR 3.1.x, Mirantis encourages you to upgrade to the latest supported version to ensure continued support and security updates.

MSR on Swarm one node to multi node scaling failure

The RethinkDB node can fail when scaling MSR on Swarm on a new cluster from one node to several nodes, resulting in the generation of the following error message during execution of scale command:

level=fatal msg="polling failed with 40 attempts 1s apart: service
\"msr_msr-rethinkdb\" is not yet ready"

To prevent such a failure, pre-pull RethinkDB images on all nodes. To do so, run following command on each node in a Swarm cluster:

docker pull registry.mirantis.com/rethinkdb/rethinkdb:2.4.3-mirantis-0.1.3