Horton
Menu

Re: Refactoring postmaster's code to cleanup after child exit

Heikki Linnakangas <hlinnaka@iki.fi>

View thread
From: Heikki Linnakangas <hlinnaka@iki.fi>
To: "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2024-07-29T20:16:33Z
Lists: pgsql-hackers

Attachments

On 06/07/2024 22:01, Heikki Linnakangas wrote:
> Reading through postmaster code, I spotted some refactoring 
> opportunities to make it slightly more readable.
> 
> Currently, when a child process exits, the postmaster first scans 
> through BackgroundWorkerList to see if it was a bgworker process. If not 
> found, it scans through the BackendList to see if it was a backend 
> process (which it really should be then). That feels a bit silly, 
> because every running background worker process also has an entry in 
> BackendList. There's a lot of duplication between 
> CleanupBackgroundWorker and CleanupBackend.
> 
> Before commit 8a02b3d732, we used to created Backend entries only for 
> background worker processes that connected to a database, not for other 
> background worker processes. I think that's why we have the code 
> structure we have. But now that we have a Backend entry for all bgworker 
> processes, it's more natural to have single function to deal with both 
> regular backends and bgworkers.
> 
> So I came up with the attached patches. This doesn't make any meaningful 
> user-visible changes, except for some incidental changes in log messages 
> (see commit message for details).

New patch version attached. Fixed conflicts with recent commits, no 
other changes.

-- 
Heikki Linnakangas
Neon (https://neon.tech)

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Fix test name and username used in failed connection attempts

  2. tests: Don't fail due to high default timeout in postmaster/003_start_stop

  3. tests: Fix race condition in postmaster/002_connection_limits

  4. Pass MyPMChildSlot as an explicit argument to child process

  5. Assign a child slot to every postmaster child process

  6. Kill dead-end children when there's nothing else left

  7. Replace postmaster.c's own backend type codes with BackendType

  8. Use an shmem_exit callback to remove backend from PMChildFlags on exit

  9. Add test for dead-end backends

  10. Add test for connection limits

  11. Fix garbled process name on backend crash

  12. Consolidate postmaster code to launch background processes

  13. Fix comment on processes being kept over a restart

  14. Refactor code to handle death of a backend or bgworker in postmaster

  15. Make BackgroundWorkerList doubly-linked

  16. Minor refactoring of assign_backendlist_entry()

  17. Fix outdated comment; all running bgworkers are in BackendList

  18. Allow building with MSVC and Strawberry perl

  19. Fix rare failure in LDAP tests.

  20. Allow notifications to bgworkers without database connections.