Log the reason about the startup and shutdown, refer to hadoop

Activity

Show:
maobaolong
May 9, 2017, 2:48 AM

This is helpful for determining the root cause of a process' exit. For example, if the process exited because the system administrator ran a standard "kill," you would see 'EXITING ON SIGNAL SIGTERM' in the log.

maobaolong
May 9, 2017, 2:57 AM

If you execute kill PID the log will appear at the end of the logfile as following.
2017-05-09T10:55:03.498+08:00] [ERROR] v2.hs.JobHistoryServer.error(LogAdapter.java 69) [SIGTERM handler] : RECEIVED SIGNAL 15: SIGTERM

Andrew Audibert
May 18, 2017, 7:32 PM

The current servers should only stop when killed. If an exception causes them to stop, the exception will be logged. What other forms of shutdown are you interested in?

Assignee

Unassigned

Reporter

maobaolong

Labels

Time tracking

0m

Time remaining

32h

Affects versions

Priority

Minor
Configure