fix: replicas can only be used with clustered Redis

[carodewig]

Per fred, replicas are only supported with clustered Redis (not standalone or sentinel).

In the current codebase, when fred connects to a high-availability Redis instance in GCP, it detects the presence of a replica and attempts to connect to it. However the replica is not guaranteed to be reachable, and if it's not (like in the default high-availability setup), the client will error and reconnect repeatedly.

As far as I can tell, there's no way to disable the attempted replica connection - it will occur when the cfg(replicas) feature is enabled. Fixing the errors requires two changes:

1. replica queries are removed from all non-cluster Redis calls (change originally made in fix: don't use replicas against standalone Redis instance #8656).
2. lazy_connections is only set to false when the Redis instance is a cluster. This way, when fred detects a replica on a non-cluster instance, it won't attempt to connect to it immediately, triggering the error.

NB: fred does see the high-availability Redis instance as a Centralized (aka standalone) instance, so determining whether the instance is a cluster can still happen via client_config.server.is_clustered().

---

Because of the specifics of how to encounter this issue, I haven't added any tests - just tested manually with RTF (see PR88 in the morgue).

No amendments to the Redis replicas changeset, as that already called out 'clustered' Redis specifically (not sentinel).

---

Checklist

Complete the checklist (and note appropriate exceptions) before the PR is marked ready-for-review.

• PR description explains the motivation for the change and relevant context for reviewing
• PR description links appropriate GitHub/Jira tickets (creating when necessary)
• Changeset is included for user-facing changes
• Changes are compatible¹
• Documentation² completed
• Performance impact assessed and acceptable
• Metrics and logs are added³ and documented
• Tests added and passing⁴
  • Unit tests
  • Integration tests
  • Manual tests, as necessary

Exceptions

Note any exceptions here

Notes

Footnotes

1. It may be appropriate to bring upcoming changes to the attention of other (impacted) groups. Please endeavour to do this before seeking PR approval. The mechanism for doing this will vary considerably, so use your judgement as to how and when to do this. ↩

2. Configuration is an important part of many changes. Where applicable please try to document configuration examples. ↩

3. A lot of (if not most) features benefit from built-in observability and debug-level logs. Please read this guidance on metrics best-practices. ↩

4. Tick whichever testing boxes are applicable. If you are adding Manual Tests, please document the manual testing (extensively) in the Exceptions. ↩

[github-actions]

@carodewig, please consider creating a changeset entry in /.changesets/. These instructions describe the process and tooling.

[apollo-librarian]

✅ Docs preview has no changes

The preview was not built because there were no changes.

Build ID: 3442318b879fcdf40b90b87e
Build Logs: View logs

[carodewig]

FYI: this PR musg be true-merged to avoid conflicts due to the cherry-picked commit.

[bnjjj]

That sounds painful...