Thanks a lot for digging into it.
See here my work flow:
I have two jar files: one with my code (A) and one with all dependencies
(B). The one with the code is rebuilt regularly.
1. build jar file A v1
2. start zeppelin
3. remove dependencies for A and B in the spark interpreter, save, and add
them again (this seems to reset zeppelin to the correct state again)
4. run the notebook => result shows code for A v1 is used => OK
5. build jar file A v2 and replace A v1 in the location specified in the
interpreter dependencies for A
6. restart the interpreter
7. run the notebook => result shows code for A v1 is still used => NOK
8. restart interpreter
9. copy jar file A v2 to localRepo (replacing A v1)
10. run the notebook => result shows code for A v2 is now used => OK
I kept a copy of the log file, happy to share it with you outside the
View this message in context:
Sent from the Apache Zeppelin Users (incubating) mailing list mailing list
archive at Nabble.com.