The cluster doesn't start correctly

[rkudryashov]

My Environment

• ArangoDB Version: 3.11.11
• Deployment Mode: Cluster
• Deployment Strategy: Docker Compose
• Configuration: see compose.yaml
• Infrastructure: own
• Operating System: Windows 11
• Total RAM in your machine: 64Gb
• Disks in use: SSD
• Used Package: Docker image

Steps to reproduce

1. Clone the repro-project
2. Execute docker compose up
3. See logs

Problem:

It seems that with the specified ArangoDB version, the cluster doesn't start correctly, that is, coordinators print multiple errors:

coordinator1  | 2024-12-03T13:04:42Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'
coordinator2  | 2024-12-03T13:04:43Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'
coordinator2  | 2024-12-03T13:04:43Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'
coordinator3  | 2024-12-03T13:04:43Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'
coordinator3  | 2024-12-03T13:04:43Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'
coordinator1  | 2024-12-03T13:04:43Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'
coordinator1  | 2024-12-03T13:04:43Z [1] ERROR [f29ef] {communication} did not find endpoint of server 'PRMR-083fcb5f-4fbf-4952-b8ae-b6e8ac9a3e1b'

Also, one of DBs prints errors:

db3           | 2024-12-03T13:04:52Z [1] ERROR [77a84] {communication} did not understand destination ''
db3           | 2024-12-03T13:04:52Z [1] WARNING [1254a] {maintenance} SynchronizeShard failed to get a count on leader _system/s22: A cluster backend which was required for the operation could not be reached
db3           | 2024-12-03T13:04:52Z [1] ERROR [77a84] {communication} did not understand destination ''
db3           | 2024-12-03T13:04:52Z [1] WARNING [1254a] {maintenance} SynchronizeShard failed to get a count on leader _system/s24: A cluster backend which was required for the operation could not be reached
db3           | 2024-12-03T13:04:52Z [1] ERROR [77a84] {communication} did not understand destination ''
db3           | 2024-12-03T13:04:52Z [1] WARNING [1254a] {maintenance} SynchronizeShard failed to get a count on leader _system/s26: A cluster backend which was required for the operation could not be reached
db3           | 2024-12-03T13:04:52Z [1] ERROR [77a84] {communication} did not understand destination ''
db3           | 2024-12-03T13:04:52Z [1] WARNING [1254a] {maintenance} SynchronizeShard failed to get a count on leader _system/s25: A cluster backend which was required for the operation could not be reached

Expected result:

No errors in the logs.

Additional context:

There is no a single error in the logs, if I downgrade the version to 3.11.10. Also, there is no error, if I don't change the version and do remove definitions of coordinators from Docker Compose file.