Keep current sync rule storage active when restarting replication due to errors

[rkistner]

There are some scenarios that trigger a restart of replication:

1. On Postgres, when the logical replication slot is invalid or deleted.
2. On MongoDB, when the resumeToken is not valid anymore.

In those cases, replication is restarted by creating a new sync-rules copy. The issue is that the active storage was immediately stopped, which could result in an unexpected [PSYNC_S2302] No sync rules available error.

This change instead keeps the current sync-rules storage active, which will be stopped automatically once the new storage has finished initial replication. This means clients can still continue syncing the (stale) data while re-replicating.

To implement this, a new ERRORED state is introduced for sync rules. This is the same as ACTIVE in terms of serving sync requests, but is not replicated.

This fixes some other issues:

1. For Postgres source db, when a replication slot is removed completely, handle it the same as an invalid slot. Previously, it would clear storage and re-create the same slot, which causes issues if that is the active sync rules storage.
2. When clearing storage for old sync rules, don't fire storageCreated hooks. Those caused issues in some cases if the sync rules had a parse error.

[changeset-bot]

🦋 Changeset detected

Latest commit: 827da3e

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 9 packages

Name	Type
@powersync/service-module-postgres-storage	Patch
@powersync/service-module-mongodb-storage	Patch
@powersync/service-module-postgres	Patch
@powersync/service-module-mongodb	Patch
@powersync/service-core	Patch
@powersync/service-module-mysql	Patch
@powersync/service-image	Patch
@powersync/service-core-tests	Patch
test-client	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR