VPS graphs are not displayed in SolusVM 2

Symptoms 

VPS graphs are not displayed in SolusVM 2

Cause

Rabbitmq queue responsible for VPS metrics is corrupted and not running:

docker exec -ti `docker ps | grep rabbitmq | awk {'print $1'}` rabbitmq-queues quorum_status vm-metrics-v2
Status of quorum queue vm-metrics-v2 on node rabbit@rabbitmq310 ...
┌────────────────────┬────────────┬───────────┬──────────────┬────────────────┬──────┬─────────────────┐
│ Node Name          │ Raft State │ Log Index │ Commit Index │ Snapshot Index │ Term │ Machine Version │
├────────────────────┼────────────┼───────────┼──────────────┼────────────────┼──────┼─────────────────┤
│ rabbit@rabbitmq310 │ noproc     │           │              │                │      │                 │
└────────────────────┴────────────┴───────────┴──────────────┴────────────────┴──────┴─────────────────┘

Resolution

1. Access SolusVM 2 Management node via SSH
2. Stop the worker responsible for vm-metrics queue:

   # docker exec -ti `docker ps | grep worker | awk {'print $1'}` supervisorctl stop rabbitmq-worker-vm-metrics-v2:rabbitmq-worker-vm-metrics-v2_00

3. Stop rabbitmq service:

   # docker service scale solus_rabbitmq=0

4. Move the directory containing the corrupted queue files:

   # mv /usr/local/solus/rabbitmq/mnesia ~/

5. Start the rabbitmq service:

   # docker service scale solus_rabbitmq=1

6. Start the worker from step 2:

   # docker exec -ti `docker ps | grep worker | awk {'print $1'}` supervisorctl start rabbitmq-worker-vm-metrics-v2:rabbitmq-worker-vm-metrics-v2_00