openai / openai/codex

Repeated failures amplify unbounded span and analytics logging

Open
#43,768 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug performance
Dominant language
Rust
Stars
125k
Forks
19.4k
PR merge metrics
PR metrics pending

Description

Repeated failures amplify unbounded span and analytics logging

  1. Trigger a persistent history-projection error while multiple tasks are loaded.
  2. Leave the tasks idle for 30 minutes.
  3. Count identical warning bodies and inspect the analytics queue.

Expected: identical failures are rate-limited with a suppressed-count summary; span fields are size-bounded and secrets are redacted before queueing.

Actual: the same projection failure is logged on every retry. In one 24-hour sample, 5,635 projection warnings coincided with 172 analytics queue full events and 2,876 discarded analytics events. Suggested fix: keyed exponential log suppression, bounded/redacted fields before serialization, and one terminal summary carrying the suppressed count.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by tracing the persistent history-projection failure through its retry logging, span-field serialization, and analytics queueing paths. Reproduce the repeated-failure scenario with multiple loaded tasks and inspect warning counts and queue behavior. Done means keyed failures are suppressed with a terminal count, serialized span fields are bounded and redacted, and the analytics queue no longer receives unbounded duplicate events.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
analytics, observability
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.