🛠️Backstage 1.55 Adds Auto-Recovery for Crashed Tasks
Auto-recovery for crashed tasks in Backstage
TL;DR
Backstage 1.55 introduces automatic recovery for tasks that crash or time out, reducing downtime. The new release also drops deprecated APIs and adds a 50 MB limit for the new database provider.
Backstage 1.55 introduces automatic recovery for tasks that crash or time out, a long-awaited feature for developers. This means tasks can resume from the last completed step, improving reliability and reducing manual intervention. The release also removes deprecated catalog APIs and introduces a new database provider with a 50 MB limit for development use. For production, teams are advised to use the GCP module or similar external stores. TechDocs now strips MkDocs plugins not on a built-in allowlist, and permission policies must explicitly allow scaffolder.template.dry-run. The release includes three patch updates since September 15, fixing issues and adding security improvements.
Key Points
Backstage 1.55 adds auto-recovery for tasks that crash or time out, resuming from the last completed step.
Deprecated catalog APIs are removed, and a new database provider with a 50 MB limit is introduced for development.
Production installations should use the GCP module or similar external stores for better performance and reliability.
TechDocs now strips MkDocs plugins not on a built-in allowlist, requiring extra plugins to be listed under techdocs.generator.mkdocs.dangerouslyAllowAdditionalPlugins.
Three patch releases since September 15 fix issues and add security improvements, including fixes in seven packages.
Why It Matters
If you're using Backstage for CI/CD, the auto-recovery feature in 1.55 can significantly reduce downtime and improve reliability. Teams should also be aware of the new security fixes and the changes to the database provider and TechDocs plugins.
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.
Join 3,568 builders reading daily.