Vladimir Ozerov as far as I recall you have faced with the issue when a slow 
client affected performance of the whole cluster recently. Please chime in this 
discussion confirming that the symptoms are the same. Most likely you already 
created a JIRA ticket for this issue.

> Also it's not clear why the nodes would require to GC because all the caches
> are held off-heap and we have  a 10G heap running G1GC.

Your logic can fill up Java heap with temporal objects. Every time you need to 
get a value from cache its copy is moved from off_heap to Java heap for a 
period of time until your code holds a reference to it. This is related to all 
kind of queries. You need to get heap dumps and/or heap histogram to see what 
kind of objects are in your heap.

—
Denis

> On Sep 16, 2016, at 6:47 AM, yfernando <yohan.ferna...@tudor.com> wrote:
> 
> Thanks for your reply Anmol. Do you know if there is a bug logged against
> this which we can track?
> 
> Also it's not clear why the nodes would require to GC because all the caches
> are held off-heap and we have  a 10G heap running G1GC.
> 
> 
> 
> --
> View this message in context: 
> http://apache-ignite-users.70518.x6.nabble.com/One-failing-node-stalling-the-whole-cluster-tp5372p7799.html
> Sent from the Apache Ignite Users mailing list archive at Nabble.com.

Reply via email to