Hi Don,
Thank you for your suggestion. Because the log file is more than 5GB on the production system, so we cannot track the detail log. Actually we don’t know what the exact time point was.
Base on the API document, the default value of waitForIDR is 1(true). When true, new files will only be created on IDR boundaries. So i think the parameter is not the caused reason. But thank you any way.
Wish more discussion with the guys who encountered the same issue.