Merge pull request #638 from dbaggerman/fix-stuck-instances

Power off immediately if cloud-init fails
buildkite · Oct 23, 2019 · 70a6590 · 70a6590
2 parents db97804 + 86eb2a8
commit 70a6590
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 0 deletions.
diff --git a/packer/linux/conf/buildkite-agent/systemd/cloud-final.service.d/10-power-off-on-failure.conf b/packer/linux/conf/buildkite-agent/systemd/cloud-final.service.d/10-power-off-on-failure.conf
@@ -0,0 +1,3 @@
+[Unit]
+OnFailure=poweroff.target
+OnFailureJobMode=replace-irreversibly
diff --git a/packer/linux/scripts/install-buildkite-agent.sh b/packer/linux/scripts/install-buildkite-agent.sh
@@ -54,6 +54,9 @@ sudo chown -R buildkite-agent: /var/lib/buildkite-agent/plugins
 echo "Adding systemd service template..."
 sudo cp /tmp/conf/buildkite-agent/systemd/buildkite-agent.service /etc/systemd/system/buildkite-agent.service
 
+echo "Adding cloud-init failure safety check..."
+sudo cp /tmp/conf/buildkite-agent/systemd/system/cloud-final.service.d/10-power-off-on-failure.conf /etc/systemd/system/cloud-final.service.d/10-power-off-on-failure.conf
+
 echo "Adding termination scripts..."
 sudo cp /tmp/conf/buildkite-agent/scripts/stop-agent-gracefully /usr/local/bin/stop-agent-gracefully
 sudo cp /tmp/conf/buildkite-agent/scripts/terminate-instance /usr/local/bin/terminate-instance