403800be22
- 6 deterministic snapshot tests at 120x40 with FakeClock frozen at 2026-01-15T12:00:00Z - Buffer-to-plaintext serializer resolving chars, graphemes, and wide-char continuations - Golden file management with UPDATE_SNAPSHOTS=1 env var for regeneration - Snapshot diff output on mismatch for easy debugging - Tests: dashboard, issue list, issue detail, MR list, search results, empty state - TERMINAL_COMPAT.md template for manual QA across iTerm2/tmux/Alacritty/kitty/WezTerm
41 lines
1.3 KiB
Plaintext
41 lines
1.3 KiB
Plaintext
Dashboard > Issues > Issue
|
|
#101 Add retry logic for transient failures
|
|
opened | alice | backend, reliability | -> bob, carol
|
|
Milestone: v2.0 | Due: 2026-02-01
|
|
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
|
|
## Problem
|
|
|
|
Transient network failures cause cascading errors in the ingestion pipeline. We need exponential backoff with jitter.
|
|
|
|
## Approach
|
|
|
|
1. Wrap HTTP calls in a retry decorator
|
|
2. Use exponential backoff (base 1s, max 30s)
|
|
3. Add jitter to prevent thundering herd
|
|
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
|
|
Discussions (0)
|
|
Loading discussions...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
NORMAL q:quit esc:back ?:help C-p:palette o:browser P:scope gh:home gi:issues gm:mrs g/:search gt:timeline
|