I have faced exact issue, when trying to import around 2 TB data into HBase. There are following ways which can solve the issue. 1. Increase hbase.hregion.memstore.block.multiplier = 8 2. increase % of RegionServer Allocated to Write Buffers from 40% to 60%. 3. Pre Split Hbase Table using start key of same Table that's might be exist on another cluster using below command. create '<HbaseTableName>',{ NAME => '<ColumnFamily>', COMPRESSION => '<Compression>'}, SPLITS=> ['<startkey1>','<startkey2>','<startkey3>','<startkey4>','<startkey5>'] Hbase Presplit enable multiple region servers can handle writes concurrently. Note: Basically this issue is appearing due to bulk write to Hbase. ... View more