thetumbled opened a new issue, #532: URL: https://github.com/apache/pulsar-client-cpp/issues/532
### Search before asking - [x] I searched in the [issues](https://github.com/apache/pulsar-client-cpp/issues) and found nothing similar. ### Version We encountered an issue on our end, which is a problem with the C++ client. This error keeps occurring: ``` E0113 21:12:39.341508 {} 32880 33642 18446744073709551615 :28 [log] RetryableOperationCache:66 | Failed to retry lookup for get-partition-metadata-persistent://XXX: Failed to create deadline_timer: eventfd_select_interrupter: Too many open files E0113 21:12:39.341516 {} 32880 33642 18446744073709551615 :28 [log] ClientImpl:197 | Error Checking/Getting Partition Metadata while creating producer on persistent://XX -- ConnectError ``` There was a network failure before the problem occurred, but it did not affect the Pulsar service. A few hours later, a C++ task had an issue, but the overall cluster services remained available. After preliminary investigation, I roughly confirmed that the client machine ran out of file descriptors (fd), causing the client to fail to connect to the broker. We are using version **3.4.2**. Claude analyzed a timer leak bug, and it seems plausible, but I don’t know much about C++. ### Minimal reproduce step Maybe network failure. ### What did you expect to see? client work right. ### What did you see instead? client can't connect to broker. ### Anything else? Here is the possible bug that is detected by claude, which need to be verified. [TIMER_LEAK_BUG_ANALYSIS.md](https://github.com/user-attachments/files/24607250/TIMER_LEAK_BUG_ANALYSIS.md) ### Are you willing to submit a PR? - [x] I'm willing to submit a PR! -- 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]
