Add correct kafka recovery functionality - #12
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: ff49d11c79
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| if (!kafka_was_reachable) { | ||
| spdlog::info("Kafka brokers are reachable again while Zeek network analysis is running."); | ||
| kafka_was_reachable = true; | ||
| bool deployed = deployZeekctl(); |
There was a problem hiding this comment.
In the Kafka-recovery path, this deployZeekctl() runs on the monitor thread, which was created after SIGINT/SIGTERM were blocked for sigwait; PosixCommandExecutor::execute then fork/execs without resetting the signal mask. When this path starts fresh Zeek workers after an outage, those processes can inherit blocked shutdown signals unlike the initial deploy, so a later Docker stop/zeekctl stop may not terminate them cleanly. Reset or unblock the signal mask before spawning zeekctl from this thread.
Useful? React with 👍 / 👎.
1 participant
No description provided.