[shrima-cf]

We should break output gates and reset actors in the event that we catch the following exceptions because they implicitly roll back any transaction:

SQLITE_FULL: database or disk full
SQLITE_IOERR: disk I/O error
SQLITE_BUSY: database in use by another process
SQLITE_NOMEM: out of memory
https://www.sqlite.org/lang_transaction.html#response_to_errors_within_a_transaction

[shrima-cf]

See #3832 for context

[github-actions]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[kentonv]

OK, at this point I just have straightforward comments. Yay!

But, what about an end-to-end test? I don't see any here nor in the internal repo. Can we test this somehow? Maybe using an SRS testDirective to inject an error?

[kentonv]

LGTM, just needs the fixups squashed and the internal test.

[github-actions]

The generated output of @cloudflare/workers-types matches the snapshot in types/generated-snapshot 🎉

[shrima-cf]

@cloudflare/wrangler Could someone help review this PR so I can merge it?

[kentonv]

I bypassed codeowner check and merged.

@danlapid why are our merges being blocked on review from the wrangler team?

[danlapid]

I bypassed codeowner check and merged.

@danlapid why are our merges being blocked on review from the wrangler team?

As a follow up to an incident of D1 a few weeks ago we set wrangler as a required CODEOWNER on types changes.

The ANT team has historically been the one to maintain and own the types generation and the generated types and they have a better grasp of the possible implications of those changes.

They are usually very quick to review but I suspect that due to Friday and Monday being bank holidays in the UK this pull request didn't get much attention.