You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
ansible.nodejs.org (IBM host infra-ibm-ubuntu2004-x64-1, 169.60.150.91) is an unmanaged, unwatched AWX (open-source Ansible Tower) instance that has been effectively dead for years and should be decommissioned rather than repaired.
What's running there
AWX 17.1.0 (ansible/awx:17.1.0), deployed via docker-compose, containers created ~5 years ago.
awx_web, awx_task, awx_redis containers, plus awx_postgres.
No nginx/reverse proxy — the AWX web container binds host ports 80/443 directly.
Key detail: it's been idle for years, then crashed
The last time this AWX instance did anything at all was around 2023-09-29 (a project sync attempt), and the last time there's evidence of actually running a job against current code was back when the checkout was at the Feb 2022 commit. Either way, it's been sitting completely idle for roughly 2.5+ years before postgres even died in Dec 2025 — the disk-full crash-loop just finished off something that was already effectively unused.
Supporting evidence:
awx_postgres has been crash-looping since 2025-12-29 (no space left on device, 605+ restart attempts logged), because the host's root disk is 100% full (99G/99G).
The project's .git/FETCH_HEAD (updated on each pre-job sync) has an mtime of 2023-09-29 11:56 UTC, and picked up no newer commits — the last sync attempt, whatever triggered it, did nothing useful.
The host's origin TLS cert (served directly by the AWX container, not through Cloudflare) is self-signed and expired 2024-04-09. The only reason https://ansible.nodejs.org still shows a valid cert to browsers is that Cloudflare is terminating TLS at the edge with the shared *.nodejs.org wildcard cert and not validating the origin connection.
Why this should be removed, not fixed
It's a privileged automation platform (AWX manages Ansible credentials/inventory) that nobody has been watching for 2.5+ years.
Disk exhaustion took down its database in Dec 2025 and nobody noticed for ~9 months — reinforcing that it has no active owner.
The AWX release (17.1.0) is multiple years out of date.
No current build/infra workflow appears to depend on it (last real activity predates most current infra automation, which now runs through this ansible/ repo directly).
Suggested next steps
Confirm nothing currently depends on this AWX instance (credentials stored only here, scheduled jobs, webhooks pointing at it, etc.)
Decommission the host (infra-ibm-ubuntu2004-x64-1 / ansible.nodejs.org, IBM, 169.60.150.91)
Remove its DNS record and Cloudflare configuration
Remove/rotate any credentials that were only ever exposed to this box (e.g. awx_password in the secrets repo)
Remove references to it from ansible/inventory.yml and ~/.ssh/config conventions once decommissioned
Summary
ansible.nodejs.org(IBM hostinfra-ibm-ubuntu2004-x64-1, 169.60.150.91) is an unmanaged, unwatched AWX (open-source Ansible Tower) instance that has been effectively dead for years and should be decommissioned rather than repaired.What's running there
ansible/awx:17.1.0), deployed via docker-compose, containers created ~5 years ago.awx_web,awx_task,awx_rediscontainers, plusawx_postgres.Key detail: it's been idle for years, then crashed
The last time this AWX instance did anything at all was around 2023-09-29 (a project sync attempt), and the last time there's evidence of actually running a job against current code was back when the checkout was at the Feb 2022 commit. Either way, it's been sitting completely idle for roughly 2.5+ years before postgres even died in Dec 2025 — the disk-full crash-loop just finished off something that was already effectively unused.
Supporting evidence:
awx_postgreshas been crash-looping since 2025-12-29 (no space left on device, 605+ restart attempts logged), because the host's root disk is 100% full (99G/99G).nodejs/buildunder/var/lib/awx/projectsis pinned at commit472b2953("Update README.md", doc: s/Node.js Foundation/Node.js Project/g #2874, merged 2022-02-21)..git/FETCH_HEAD(updated on each pre-job sync) has an mtime of 2023-09-29 11:56 UTC, and picked up no newer commits — the last sync attempt, whatever triggered it, did nothing useful.https://ansible.nodejs.orgstill shows a valid cert to browsers is that Cloudflare is terminating TLS at the edge with the shared*.nodejs.orgwildcard cert and not validating the origin connection.Why this should be removed, not fixed
ansible/repo directly).Suggested next steps
infra-ibm-ubuntu2004-x64-1/ansible.nodejs.org, IBM, 169.60.150.91)awx_passwordin the secrets repo)ansible/inventory.ymland~/.ssh/configconventions once decommissioned🤖 Generated with Claude Code