Hi Travis CI Support,
All our builds have been failing for the past 2 days across multiple repositories. After analyzing the raw job logs, we have identified the exact failure point — builds are hanging during Travis’s internal infrastructure setup steps, before any of our .travis.yml configuration is executed.
Failure Pattern:
On dist: focal builds, the hang occurs at the fix_perforce_key setup event. In the passing build, this step completes and prints OK. In the stuck build, the event starts but never completes, producing zero output until Travis’s 10-minute no-output timer fires and kills the job.
On dist: jammy builds, the hang occurs immediately after update_apt_keys completes, before fix_hhvm_source begins.
In both cases, the build is terminated with:
“No output has been received in the last 10m0s”
Comparison of Job Logs from our 2 repos:
| Passing | Stuck | |
|---|---|---|
| repo_A (focal) | Job 639395966 |
Job 639441032 |
| repo_B (jammy) | Job 639432471 |
Job 639440373 |
All four jobs share:
-
Worker version:
deploy_2025.08.20.CI.1 -
focal VM image:
travis-ci-ubuntu-2004-1761808138-4ffea187-dirty -
jammy VM image:
travis-ci-ubuntu-2204-1761745415-2d890300-dirty -
travis-buildversion:2812e526
The VM images, worker versions, and build configurations are identical between passing and failing jobs. No changes were made to our CI configuration.
What we have ruled out:
-
Not a cache issue (tested with
cache: false) -
Not a dist pinning issue (both
focalandjammyare affected) -
Not application code or dependency related (hang occurs before git clone in some cases)
-
Not a Travis platform-wide outage (status page shows green)
-
Not a GCP zone issue (both passing and failing builds ran on
gce-production-1)
Could you plz help figure out why the fix_perforce_key and post-update_apt_keys setup steps are intermittently hanging on your worker infrastructure.