Follow https://community.cloudera.com/t5/Support-Questions/Hive-staging-directory-not-getting-cleaned-up/td-p/43976,if you are running large number of  queriesies through hue ... View more

executing a file with multiple queries in it should work. But each statement is executed individually and is non-atomic (they are not executed within a single transaction). Please post exceptions or errors if this has not worked for you. ... View more

This (NULL issue) can be caused by data schema mismatch. ... View more

When you are passing your truststore make sure that the truststore consist of the Client hosts that are comminucating with your HiveServer2.   It is always safe to have 10 nodes then its good to have the list of hosts(Or Aliases) trusted in your Truststore.   ... View more

we are facing another ISSSUE: inserted only partial number of rows. The source table has 7,070,150,676. The insert query ran and we could see that from the source table same number of rows were transferred and beeline end screenshot also shows numRows = 7,070,150,676 but in the target hive table only 500,912,361 many rows were inserted.     The user is using QueryGrid.   Please hekp us to find the issue. ... View more

Hi All,   Facing below error while running the hive command.   MSCK REPAIR TABLE flexdto_standin_avro;   Error while processing statement: FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask Your inputs is much appreciated.   Thanks Yasmin ... View more

SELECT * does not trigger MR job, while SELECT struct_col.field will, which means the jar is probably available on HS2 host, but not available on the NM hosts. As Naveen mentioned, the class is in hive-hcatalog-xxx.jar, so you can put this jar into Hive AUX directory on HS2 host and then it should be distributed to the cluster while query is running in MR mode. Regarding JsonSerDe in C6.1, Naveen is right and it has been confirmed. ... View more

Naveen, Thanks for exhaustive answer. I am a newbie so I might be wrong, but after some experiments I tend to believe that the current output.format.string, as it is written in tutorial is wrong.   Currently it is:   "%1$$s %2$$s %3$$s %4$$s %5$$s %6$$s %7$$s %8$$s %9$$s" I believe it should be: "%1$s %2$s %3$s %4$s %5$s %6$s %7$s %8$s %9$s" What makes me think so? I have tried, just for fun and experimenting, inserting a new row in intermediate_access_log table in hive. And the original output.format.string was making the statement to fail. After the change of the format string, the new row was nicely inserted. ... View more

In below mentioned post when I fire query like below, works fine    select * from working.ddr2_raw_actual_test where dest_icao='NULL'   But with actual file it is not working. I am getting resultset in hive which is correct but same query not able to get in impala.   Thanks and Regards, Naveen Srikanth D ... View more

I have managed to solve my problem, it was a silly little mistake that I was making. I created JSON table using: ADD JAR hdfs://hwmaster01.com/user/root/hive-serdes-1.0-SNAPSHOT.jar; CREATE TABLE tweets_pqt ( id BIGINT, created_at STRING, source STRING, favorited BOOLEAN, retweeted_status STRUCT< text:STRING, user:STRUCT<screen_name:STRING,name:STRING>, retweet_count:INT>, entities STRUCT< urls:ARRAY<STRUCT<expanded_url:STRING>>, user_mentions:ARRAY<STRUCT<screen_name:STRING,name:STRING>>, hashtags:ARRAY<STRUCT<text:STRING>>>, text STRING, user STRUCT< screen_name:STRING, name:STRING, friends_count:INT, followers_count:INT, statuses_count:INT, verified:BOOLEAN, utc_offset:INT, time_zone:STRING>, in_reply_to_screen_name STRING ) ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe' After putting that above mentioned JAR file in Cloudera Manager's "Hive Auxiliary JARs Directory" (on un-managed machine you can find that under "hive.aux.jars.path" property in hive--site.xml) Then I created the Parquet table with same structure as above with a little change:   ADD JAR hdfs://hwmaster01.com/user/root/hive-serdes-1.0-SNAPSHOT.jar; CREATE TABLE tweets_pqt ( id BIGINT, created_at STRING, source STRING, favorited BOOLEAN, retweeted_status STRUCT< text:STRING, user:STRUCT<screen_name:STRING,name:STRING>, retweet_count:INT>, entities STRUCT< urls:ARRAY<STRUCT<expanded_url:STRING>>, user_mentions:ARRAY<STRUCT<screen_name:STRING,name:STRING>>, hashtags:ARRAY<STRUCT<text:STRING>>>, text STRING, user STRUCT< screen_name:STRING, name:STRING, friends_count:INT, followers_count:INT, statuses_count:INT, verified:BOOLEAN, utc_offset:INT, time_zone:STRING>, in_reply_to_screen_name STRING ) --ROW FORMAT SERDE 'org.apache.hive.hcatalog.data.JsonSerDe' STORED AS PARQUET; I inserted into parquet table succesfully the moment I commented that line.     ... View more