Hello, We use ignite 2.2.0 (we migrated from 2.1.0) We have 3 Ignite nodes, each with 32GB of memory. We are loading on a node a large amount of data (aprox. 5GB) on one node. The chunkSize is calculated at 1.94GB. Sometimes we get Connection reset by peer on a node and the caches are closed afterwards. Can this be due to the large chunkSize? (like node1 is trying to send to node2 1.94gb of data?) If yes, is there a way to fix that, like setting a smaller chunkSize? We tried with memory policies, but the chunkSize is always calculated. On ignite 2.1.0 the chunkSize was calculated at 1.34GB and we didn't have any issue. Thank you.
-- Sent from: http://apache-ignite-users.70518.x6.nabble.com/
