Croway commented on PR #27958: URL: https://github.com/apache/camel/pull/27958#issuecomment-6233843459
Follow-ups are in f5e8ae9c6d94 (rebased on main again, the upgrade guide conflict kept both sides, hence the force push), and the title no longer says "timeouts are errors". @allthingssecurity thanks, the lag is real and the fix follows your shape: the `total` of the receive status before enabling bounds this call's messages (`uid` above it), one call at a time per pid, and the status read after the stop gives the exact count. I went one step further than counting: the call waits (3 s at most) until the status shows the console disabled and the file holds the messages the count announces, so the messages taken between the dump it read and the stop are returned too, not only counted. On the timeout path that turns "received 0" into the real number. The wait is bounded, and the answer says when the process did not confirm the stop or had not written every message in time. Both doc pages describe `received` as every message the consumer took and the one-call-at-a-time rule. Tests: camel-jbang-mcp test package 512, camel-jbang-core runtime tests, TUI `AiPanelPromptBudgetTest` (no change), all passing. _Claude Code on behalf of Croway_ -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
