cdap.sh restart should kill hanging processes after a timeout
Description
Release Notes
Activity
Sreevatsan Raman February 10, 2017 at 12:10 AMEdited
We are not seeing problems after removing kafka and zookeeper. So there is no platform issue. Marking this as resolved. If there are any specfic app related issue, we should open separate JIRAs to fix as killing process for convenience only masks the underlying issue.
Ali Anwar February 7, 2017 at 10:48 PM
Discussed with Russ, he was running a Hydrator pipeline that was loading >1GB of data into CDAP standalone, and wasn't able to stop the MR job from the UI.
Russ Savage January 31, 2017 at 6:06 PM
Assigning to for reassignment.
Russ Savage January 28, 2017 at 7:43 PM
i still see a lot of dots sometimes, although i think it is related to a program, not zookeeper. Like, the below example where it took over 20 mins to stop.
Priyanka Nambiar January 26, 2017 at 11:07 PM
audit logs have been moved to transaction message, and this may have appeased this situation. Russ, could you take a look if this is still happening?
Sometimes, after sleeping your laptop and then trying to restart cdap, it hangs and just keeps printing periods on the screen. After a timeout, the script should use a more forceful method to kill the hanging process.