lianghuan-xatu commented on issue #5012: URL: https://github.com/apache/seatunnel/issues/5012#issuecomment-1759303154
> Hi, I'm leaving a question in passing. When, main node crash I think connection between new selected node and file system that save jar package is important. Am I right? and doesn't this change make recover time from abnormal master node long? Thank you very much for raising this question and your viewpoint is correct. The previous solution was to issue a pull request for missing Jar packets from the slave node to the master node, which would result in a longer recovery time for abnormal nodes. The slave node needs to download all missing Jar packet files from the master node to complete node recovery. This pull method also requires the service to rely on a distributed system, so I did not use pull in the future. Instead, I pushed the Jar package files required for task execution through the master node to each slave node for storage, so that each node stored a copy of the Jar package file, thereby avoiding the transmission delay required for Jar package file recovery between nodes. -- 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]
