[NCBC-121] Null Reference Exception during rebalance Created: 17/Sep/12  Updated: 03/Jan/13  Resolved: 21/Nov/12

Status: Closed
Project: Couchbase .NET client library
Component/s: None
Affects Version/s: 1.1.6
Fix Version/s: 1.2 Beta-2

Type: Bug Priority: Blocker
Reporter: Mark Nunberg Assignee: John Zablocki (Inactive)
Resolution: Fixed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: PNG File net-bug.png    

 Description   
Happens during the rebalancing-out of two nodes.

Reproduce:

* Perform get/set operations continuously
* Rebalance out two nodes
poof!

I was using multiple client handles for this test.. it seems each of them got that error.

 Comments   
Comment by Farshid Ghods (Inactive) [ 17/Sep/12 ]
mark,

Just wondering if major is the right priority for this issue given that this happens during rebalancing
Do users tend to ignore one or two errors and are okay with retries for these type of errors
Comment by Mark Nunberg [ 17/Sep/12 ]
This causes a complete application crash.
Comment by Mark Nunberg [ 19/Sep/12 ]
I've patched up Enyim a bit so that it no longer crashes -- the problem being some apparent reentrancy and race condition issues when a node is dying.

I'll file a gerrit patch when I figure out how to use this thing on my windows box
Comment by Mark Nunberg [ 27/Sep/12 ]
Interim changeset is here
https://github.com/mnunberg/EnyimMemcached/commit/f2b161a0909badf7981e055605d3a20f9a5db4dd
Comment by Matt Ingenthron [ 04/Oct/12 ]
John, can you evaluate the severity of this? This has been a blocker for QE tests.
Comment by John Zablocki (Inactive) [ 23/Oct/12 ]
Merged Mark's code into our Enyim fork
Comment by John Zablocki (Inactive) [ 21/Nov/12 ]
Published
Generated at Wed Oct 01 06:23:07 CDT 2014 using JIRA 5.2.4#845-sha1:c9f4cc41abe72fb236945343a1f485c2c844dac9.