liangrui198 opened a new issue, #3514: URL: https://github.com/apache/celeborn/issues/3514
When Celeborn was discovered to be using Spark Shuffle, TCL was operating at 10K - 15K, with disk ioutil reaching 100%. Even when the disk's blkio.throttle.read_bps_device was limited to 80%, the disk ioutil usage still remained at 100%. This indicates that the data read is very small, but the IOPS is high. Are there any optimization methods? -- 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]
