Hi Ray, Thank you for thread dumps. 'Failed to wait for partition map exchange' is related to rebalancing. What can be reasons of rebalancing, is it possible some nodes were joining or left topology? Data load itself can't cause rebalancing, partitions are not moved if cluster is stable.
If Persistent Data store mode enabled then algorithm used for eviction is Random-LRU (most recent access timestamp is stored for each data page). Additional configuration is not required. SQL index is also allocated in the default memory region (or cache related region if configured). Because slow HDD is used, increasing available memory will definitely help with upload speed. Sincerely, Dmitriy Pavlov ср, 18 окт. 2017 г. в 9:17, Ray <[email protected]>: > After another try to ingest the data, the job stuck after about 10 minutes. > > And the log is filled with these messages > > [2017-10-18T06:14:27,706][INFO > ][grid-timeout-worker-#63%null%][IgniteKernal] FreeList [name=null, > buckets=256, dataPages=300884, reusePages=228659] > [2017-10-18T06:14:19,556][WARN ][exchange-worker-#106%null%][diagnostic] > Failed to wait for partition map exchange [topVer=AffinityTopologyVersion > [topVer=338, minorTopVer=0], node=a29d9d3b-19d1-4279-942e-28763ca82e3a]. > Dumping pending objects that might be the cause: > > And still no "Page evictions started, this will affect storage performance" > message in the log. > > I'm guessing is it that I didn't specify which eviction policy to use for > ignite server, so ignite didn't evict the page to disk? > > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ >
