CODEWITHSUNDEEP
hadoophbasebulk-load
pyOwner
pyOwner

Reputation: 650

hbase completebulkload long running

A hadoop cluster was upgraded from CDH 4.4 to CDH 5.7.6, HBASE: 0.94 -> 1.2.0 After that time of hbase completebulkload process has increased from 2 hours to 3-4 for small tables and 10+ hours for big tables. Why it happens? What should I check for?

Upvotes: 0

Views: 125

Answers (1)

pyOwner
pyOwner

Reputation: 650

The problem was in disabled blockcache. If you load hfiles through the code using LoadIncrementalHFiles it works properly. https://issues.apache.org/jira/browse/HBASE-10500

Upvotes: 0

Related Questions