[GitHub] [helix] esarkissian-rythmos opened a new issue #609: OnlineOffline model only works up to 0.6.9

[GitBox <gi...@apache.org>]

esarkissian-rythmos opened a new issue #609: OnlineOffline model only works up to 0.6.9
URL: https://github.com/apache/helix/issues/609
 
 
   We have been working with Helix for some time now, and we observed that in any version higher than 0.6.9, when using the simple OnlineOffline model, we observe the following:
   
   - Partitions 100
   - Single resource - e.g. buffer.queue
   - Replicas: 1
   
   Using the cli tools in the Helix distro, we saw that the cluster has the following issues:
   
   - More than one participant receives ONLINE for the same partition
   - The cluster can't settle as we see participants fighting over the same partition, in a loop, going ONLINE/OFFLINE.
   
   The only thing that seems to aleviate this problem (still testing), is to switch to 0.6.9. Any version above that has the above problems, it would appear.
   
   We tested this with a 3-node ZK 3.5.6 instance, setup locally. But it happens on our AWS setups too. It just seems like in the latest versions the OnlineOffline model at leats is completely unusbale.
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@helix.apache.org
For additional commands, e-mail: reviews-help@helix.apache.org