Skip to content

Make SQL read/write retry settings configurable #108

Description

@StuartFerguson

Problem

This service uses SQL Server for both read and write paths, so a transient database failure can affect either side of the request flow.

Related investigation: TransactionProcessor#1831

Requested change

Add configurable SQL retry settings for both read and write operations.

What should be configurable

  • max retry count
  • max retry delay
  • any additional transient error numbers if needed

Notes

  • Keep read and write retry settings separate if this service needs different policies.
  • Use sensible defaults so existing behavior does not regress.
  • If any code runs migrations or grouped database work, make sure it still runs correctly under the retry strategy.

Acceptance criteria

  • Retry settings can be adjusted from configuration
  • Both read and write paths use the configured policy
  • Existing behavior is preserved by default where possible
  • Add or update tests if the service has coverage around database access
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions