Assign / Unassign issue

[ArtemisMucaj]

Hi,

I've been trying to resolve an issue when adding consumer scripts with the rebalance callback.

1. Run X instances of the Consumer
2. Run another Consumer
3. Wait for rebalance and repeat 2.
4. From time to time one of the Consumers can't rebalance and it starts receiving {"message":"Local: Erroneous state","code":-172,"errno":-172,"origin":"kafka"} errors

This is what I get when running in gdb.

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff7fe7780 (LWP 22887)]
0x00007ffff490b884 in NodeKafka::Consumer::NodeGetAssignments (info=...) at ../src/consumer.cc:493
493 Nan::Newv8::String(part->topic()).ToLocalChecked());

(I'm not entirely sure this is the reason)

[ArtemisMucaj]

After further debugging I still don't really know what the core of the issue is (but I doubt its the segfault).
The one thing I keep observing is that sometimes "assign" method is not called and the Consumer can never recover.

[ArtemisMucaj]

I figured out what causes the issue.

https://github.com/Blizzard/node-rdkafka/blob/master/src/consumer.cc#L156
https://github.com/Blizzard/node-rdkafka/blob/master/src/consumer.cc#L175

Changing m_is_subscribed state on assign and unassign doesn't make sense here as it causes RdKafka::ERR__STATE error when consuming even though librdkafka already maintains a subscribed / unsubscribed state internally.

Will write a PR asap.

[webmakersteve]

Merged the PR so I believe the issue is closed!

As far as I can tell, it needs to be included in the next minor release because it can theoretically be a breaking change if someone relied on the throwing behavior in their application.