towerops

graham/towerops

Fork 0

Commit graph

Author	SHA1	Message	Date
Graham McIntire	532d88ffb9	perf: disable Req retry for faster error tests - Add retry: false to VISP Client HTTP requests - Add retry: false to ReleaseChecker GitHub API requests - Remove unused Plug.Conn import from RemoteIpLogger - Remove unused default parameter from req_get/2 Results: - VISP sync test: 7007ms → 113ms (62x faster) - ReleaseChecker test: 7004ms → 17ms (402x faster) - UserResetPasswordLive test: 1495ms → 284ms (5x faster) Req's default retry behavior (1s, 2s, 4s exponential backoff) was causing 7-second delays for HTTP 500/503 error responses in tests. For these clients, immediate failure is preferred over retries.	2026-03-10 16:19:31 -05:00
Graham McIntire	de986bddf6	Prevent Oban polling/monitoring job stacking per device When the Oban queue backs up, the 60-second uniqueness window expires and duplicate jobs stack up per device. Switch to period: :infinity so only one poll/monitor job exists per device at any time. Add replace option to supersede stale scheduled jobs with updated scheduled_at. Remove :executing from unique states so self-scheduling works while the current job runs. Set max_attempts: 1 since retrying stale polls is pointless. Also fix all credo --strict issues across the codebase.	2026-02-16 16:37:48 -06:00
Graham McIntie	2735c1a032	Add onboarding flow for new organizations	2026-02-16 10:14:45 -06:00

Author

SHA1

Message

Date

Graham McIntire

532d88ffb9

perf: disable Req retry for faster error tests

- Add retry: false to VISP Client HTTP requests
- Add retry: false to ReleaseChecker GitHub API requests
- Remove unused Plug.Conn import from RemoteIpLogger
- Remove unused default parameter from req_get/2

Results:
- VISP sync test: 7007ms → 113ms (62x faster)
- ReleaseChecker test: 7004ms → 17ms (402x faster)
- UserResetPasswordLive test: 1495ms → 284ms (5x faster)

Req's default retry behavior (1s, 2s, 4s exponential backoff) was
causing 7-second delays for HTTP 500/503 error responses in tests.
For these clients, immediate failure is preferred over retries.

2026-03-10 16:19:31 -05:00

Graham McIntire

de986bddf6

Prevent Oban polling/monitoring job stacking per device

When the Oban queue backs up, the 60-second uniqueness window expires
and duplicate jobs stack up per device. Switch to period: :infinity so
only one poll/monitor job exists per device at any time. Add replace
option to supersede stale scheduled jobs with updated scheduled_at.
Remove :executing from unique states so self-scheduling works while
the current job runs. Set max_attempts: 1 since retrying stale polls
is pointless.

Also fix all credo --strict issues across the codebase.

2026-02-16 16:37:48 -06:00

Graham McIntie

2735c1a032

Add onboarding flow for new organizations

2026-02-16 10:14:45 -06:00

3 commits