Support Forums : DataStax

DataStax Support Forums » DataStax Enterprise

How do I update Hive table when columns are added to Cassandra?

(4 posts) (3 voices)

Started 10 months ago by jonusko
Latest reply from alexliu
This topic is not a support question

Tags:

No tags yet.

jonusko
Member

If I define all the columns for a column family in Cassandra, I see that they are automatically defined in Hive. If I use ALTER TABLE to add a column in Cassandra, this column does not get added to the Hive metastore thus it is not visible to Hive. Hive does not let me modify an external table. So how do I get my Hive table back in sync with the Cassandra column family?

Steps to reproduce:
1) Create table in cqlsh.
2) Insert row in table using cqlsh.
3) Select row in Hive to verify the data.
4) Alter table in cqlsh and add a new column.
5) Update new column with a value in cqlsh.
6) Select row in Hive to verify the new column's data.

In step #6, select * does not return the new column and select newColumn returns an error.

Posted 10 months ago #
tjake
Moderator

Did you restart Hive before #6?

Posted 10 months ago #
jonusko
Member

Yes. I was starting/stopping cqlsh and dse hive from the same shell prompt and never had them running at the same time.

I tried a lot of things to make this work, one of them was to delete the Hive database definition in the HiveMetaStore.MetaStore. I only deleted the one row in the MetaStore that corresponded with my Hive database. Now, Hive will no longer run properly. When I try to "use" any database in Hive, I get an OutOfMemoryError: GC overhead limit reached. I tried restarting dse cassandra and dse hive, but Hive still is broken. Is there anyway to fix Hive?

Posted 10 months ago #
alexliu
Member

Drop HiveMetaStore keyspace and MetaStore column family. Then restart DSE, restart HIVE, you should have a fresh start.

Posted 10 months ago #

RSS feed for this topic

Reply

You must log in to post.