The disk space occupied by a deleted row is only reclaimable via compaction and  given you have deleted some data and if the space is not reclaimed then probably you are hitting the bug  https://issues.apache.org/jira/browse/KUDU-1625   The jira stands unresolved. However if the goal is to delete the data and reclaim disk space, then you can drop partition (if range partition) in order to reclaim space.    Tombstone tablets have all their data removed from disk and don't consume significant resources. These tablet are necessary for correct operation of kudu. See - https://docs.cloudera.com/runtime/7.1.0/troubleshooting-kudu/topics/kudu-tombstoned-or-stopped-tablet-replicas.html   ... View more

Hey @smdas , thanks for your feedback, i will rephrase it for better understanding. Hive table name =hive_t1 External hive hbase table name= hb_t2 They both have identical data as per now 100 rows each Senario 1: Select t1.c1,t1.c2 from hive_t1 t1 left join(select KEY, c1 from hb_t2 where KEY like 'abc%') t2on (t1.c1=t2.c1); Here the values which i get from hbase external table are null , but expected results should be having matching values.  When i was analysing found these results also. Senario 2: select KEY, c1 from hb_t2 where KEY like 'abc%'; => returns 0 rows or no result But if i run this  select * from hb_t2 where KEY like 'abc%'; => Then i am able to see the data all the 100 rows, not able to understand this behaviour.   -Shivam   ... View more

Hello @rajatsachan    Thanks for sharing the details into the Steps used by you to resolve the issue. This would definitely assist fellow Community Members facing similar issues. If you have no further concerns, Kindly mark the Post as Solved as well.    Thanks, Smarak ... View more

Hello @snm1523    As the Post was resolved, I am marking the Post as Solved. In future, Kindly mark the Post as Solved to ensure other Community Users can reference the Post for similar issues.   Thanks for using Cloudera Community.    - Smarak ... View more

Solved. Thank you ... View more

Hello @ShamsN    Kindly update the Post, if you have solved the issue. If you continue to face the issue, Let us know & we can assist you. We requested additional details based on your Post on 12/16.    - Smarak ... View more

Hello @Madhureddy    Thanks for using Cloudera Community. Based on the post, Table "Meterevents" was loaded with 3K records & an Insert Select Operation was performed against "events_Hbase" from "Meterevents" table. The "events_Hbase" table is showing 1200 records.    We wish to check upon the following details: 1. Connect to HBase Shell & confirm the count of "HbaseEvents" table, 2. If the count of "HbaseEvents" table is 1200, Check for the Uniqueness of the 1st Column being used as ":key" while loading the Table. It's likely the RowKey is being repeated, causing an updated Version being utilised, thereby reducing the row-count.  3. Your team can check upon the above by creating 2 Tables & insert 10 unique rows (By RowKey Column) into 1 Table with 10 rows (Having, 5 Unique RowKey Values) into the 2nd Table. Next, Create 2 Hive Table using HBaseStorageHandler & perform the Insert Select SQL. Then, Check the Row Count.    - Smarak ... View more

@smdas Thank you for the answer. ... View more

Hello @TGH    Yes, After doing any HBCK2 Changes, Restart the Service as the Components have a Cached Version of the Metadata as well. Let us know how things goes.    - Smarak ... View more