fix(traige signals): Fix rare fixability failure issue (#104438)

## PR Details
+ Issue: generate_issue_summary_only task fails ~1-2% of the time with
No issue summary found for group_id error from Seer's fixability
endpoint -
https://sentry.sentry.io/issues/7082675053/events/8d6139c6d386440ea729c788b06c3aee/?referrer=merged-item
+ Possible Root cause: Database read-after-write visibility issue
between Seer servers
- Summarize endpoint (on SEER_SUMMARIZATION_URL) commits summary to DB
and returns 200
- Fixability endpoint (on SEER_SCORING_URL) immediately queries DB but
doesn't see the committed row yet
- This happens because the two endpoints run on different servers, and
DB replication lag causes the fixability read to occur before the
summarize commit is visible
+ We see this in `generate_summary_and_run_automation` too -
https://sentry.sentry.io/issues/6792150415/?project=1&query=is%3Aunresolved%20%22generate_summary_and_run_automation%22&referrer=issue-stream.
If this works here I'll add it there too.
+ Fix: Add 3-second delay before retry so that when fixability fails due
to replication lag, the retry occurs after DB has synchronized

Author: Mihir-Mavalankar

src/sentry/tasks/autofix.py

@@ -47,7 +47,7 @@ def generate_summary_and_run_automation(group_id: int) -> None:
     name="sentry.tasks.autofix.generate_issue_summary_only",
     namespace=ingest_errors_tasks,
     processing_deadline_duration=35,
-    retry=Retry(times=1),
+    retry=Retry(times=3, delay=3, on=(Exception,)),
 )
 def generate_issue_summary_only(group_id: int) -> None:
     """