Hooks - rollback on failure

[timleslie]

If a mutation fails we need to be able to rollback any operations which happened during the hooks leading up to the failure.

We need to provide a mechanism for this rollback, but it has not yet been decided what that mechanism should be.

The initial hairy session for hook design came up with the idea of providing a addRollbackStep function which could be called from hooks, which would collect all the rollback functions to be executed on failure.

Subsequent review of this in the context of nested mutations and the difficultly of tracking rollbacks introduced the idea that perhaps a simpler design would be that each field type/field/list could have a cleanup (or similar) hook and it would be up to the system designer to keep track of what needed to be called in the case of failure.

[stale]

It looks like you haven't had a response in over 3 months. Sorry about that! We've flagged this issue for special attention. It wil be manually reviewed by maintainers, not automatically closed. If you have any additional information please leave us a comment. It really helps! Thank you for you contributions. :)

[MadeByMike]

I've updated #211 and am now closing this as a dupe