[ https://issues.apache.org/jira/browse/IGNITE-3018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15270799#comment-15270799 ]
Taras Ledkov commented on IGNITE-3018: -------------------------------------- Please take a look at the attached heatmaps. The node distributions of affinity function with MD5 hash & Wang hash with bucket based algorithm are compared for 3, 64, 100, 128, 200, ..., 600 nodes. Horizontally: node's order (primary node, backup0, backup 1); Vertically: all nodes from topology; Z-order: count of node is placed in the specified order (e.g. node is primary nodes) for all partitions. > Cache affinity calculation is slow with large nodes number > ---------------------------------------------------------- > > Key: IGNITE-3018 > URL: https://issues.apache.org/jira/browse/IGNITE-3018 > Project: Ignite > Issue Type: Bug > Components: cache > Reporter: Semen Boikov > Assignee: Taras Ledkov > Priority: Critical > Fix For: 1.6 > > Attachments: 003.png, 064.png, 100.png, 128.png, 200.png, 300.png, > 400.png, 500.png, 600.png > > > With large number of cache server nodes (> 200) RendezvousAffinityFunction > and FairAffinityFunction work pretty slow . > For RendezvousAffinityFunction.assignPartitions can take hundredes of > milliseconds, for FairAffinityFunction it can take seconds. > For RendezvousAffinityFunction most time is spent in MD5 hash calculation and > nodes list sorting. As optimization we can try to cache {partion, node} MD5 > hash or try another hash function. Also several minor optimizations are > possible (avoid unncecessary allocations, only one thread local 'get', etc). -- This message was sent by Atlassian JIRA (v6.3.4#6332)