False alarm, there are always segment owners. It's an issue with start procedure where
the cache manager has members available for segment information before Hot Rod server has
added itself to the topology cache. Thx to Will for spotting it :)
On 24 Mar 2015, at 10:52, Galder Zamarreño <galder(a)redhat.com>
wrote:
Hey Dan,
Re:
https://issues.jboss.org/browse/ISPN-5208
I've created a unit test based on Matej's feedback to replicate the issue [1].
Debugging through it, I've discovered that some topology updates are sent with
segments that have no owner, hence when the key is hashed to that segment, it finds no
owner for it.
Looking at the logs, I see:
2015-03-24 10:29:57,855 4405 TRACE [org.infinispan.topology.CacheTopology]
(remote-thread-main-NodeC-p15-t4:) Current consistent hash's routing table: 0: 0, 1:
0, 2: 0, 3: 0, 4: 0, 5: 0, 6: 0, 7: 0, 8: 0, 9: 0, 10: 0, 11: 0, 12: 0, 13: 0, 14: 0, 15:
0, 16: 0, 17: 0, 18: 0, 19: 0, 20: 0, 21: 0, 22: 0, 23: 0, 24: 0, 25: 0, 26: 0, 27: 0, 28:
0, 29: 0, 30: 1, 31: 1, 32: 1, 33: 1, 34: 1, 35: 1, 36: 1, 37: 1, 38: 1, 39: 1, 40: 1, 41:
1, 42: 1, 43: 1, 44: 1, 45: 1, 46: 1, 47: 1, 48: 1, 49: 1, 50: 1, 51: 1, 52: 1, 53: 1, 54:
1, 55: 1, 56: 1, 57: 1, 58: 1, 59: 1
If having a segment that has no owners is valid scenario, what should we do about it?
- Should the server avoid sending such topology unless all segments have owners?
- Or should the client use round-robin as fallback when segment for key has no owner?
- Or is this a bug?
Cheers,
[1]
https://gist.github.com/galderz/17f1c4f21b4e7856aa6b
--
Galder Zamarreño
galder(a)redhat.com
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev