DMS Troubleshooting

AI Tools

How to See a Censored Task Output  

Issue

Task fails with "the output has been hidden due to the fact that no_log: true was specified".

Solution

docker compose run --rm -e ANSIBLE_NO_LOG=false \ -v `pwd`/inventory:/inventory deploy-dms
Note

If a task hardcodes no_log: true, comment that line out temporarily to see the error.

--remove-orphans Warning

Issue

A warning may appear that states that orphan containers (thick-clone-worker, discovery-client) were found for this project

Solution

This warning is expected. DMS runs two compose files under one project, so each compose command reports the other file's services as orphans.

Do not run compose with --remove-orphans. It will delete the thick-clone-worker and discovery-client containers. To silence the warning, set COMPOSE_IGNORE_ORPHANS=true.

mydms Reported Unhealthy on First Deploy

Issue

Task fails with "container mydms is unhealthy"; mydms is restarting with exit 255.

Solution

  1. Check the container:

docker inspect --format '{{.State.Health.Status}}' mydms

  1. If it reports healthy, re-run the playbook. It is idempotent.

  2. If it keeps restarting, increase wait_timeout (default 120s).

Note

A stale Temporal schedule from an earlier failed run makes mydms restart until it clears itself, often just after the 120s healthcheck window expires.

SSH_KEY_PATH_IN_CONT Warning

Issue

A warning may appear that the "SSH_KEY_PATH_IN_CONT" variable is not set.

Solution

This warning is expected when using username/password SSH authentication, and can be ignored.

Ansible Pipelining Failures

Issue

A task fails at module execution with rc=1, but SSH login and /usr/bin/python3 both work when tested by hand.

Solution

In ansible.cfg:

[ssh_connection] pipelining = False
Note

Login-time output on the DMS server (MOTD, banner, or an echo in .bashrc) can corrupt the pipelined module stream.