Grokbase Groups Hive user May 2011
FAQ
Hi,



I started using Toad
for Hive,looks cool! I installed the plugins in Eclipse in windows and it
started working fine.

I have a few questions:



1. Does it work with Ubuntu-Linux?

2. Job Tracker works okay, but when I look into some
tasktracker logs, the url gets changed as hostnames of the tasktrackers instead
of IPs and so I have to manually edit the URL in the JobTracker console to have
the tasktracker logs! Is there some way I could solve this?

3. Is there a way to kill a Hive Job from the Toad console?
The Stop SQL button in the SQL Editor stops the command from Toad but do not
kill the job.



Thanks,

Mike

Search Discussions

  • Peter Hall at May 16, 2011 at 1:30 am
    Hi Mike,

    1. There's no official support for Ubuntu, however I don't know of any reason why the hive features wouldn't work. The data hub is only available as a Fedora rpm, but is not required for direct hive connections.

    2. The jobtracker uses reverse DNS to determine the addresses of the tasktrackers. If it is getting this wrong you can set slave.host.name on each tasktracker or change your dns setup so reverse dns gives you the host name you want to use.

    3. Not at the moment. I'll pass this one up the chain and see if we can add it in a future release.

    Cheers,
    Peter Hall
    Quest Software
    ________________________________
    From: Mike [nano_kollam@yahoo.com]
    Sent: Sunday, 15 May 2011 2:12
    To: user@hive.apache.org
    Subject: Using Toad for Hive


    Hi,



    I started using Toad for Hive,looks cool! I installed the plugins in Eclipse in windows and it started working fine.

    I have a few questions:



    1. Does it work with Ubuntu-Linux?

    2. Job Tracker works okay, but when I look into some tasktracker logs, the url gets changed as hostnames of the tasktrackers instead of IPs and so I have to manually edit the URL in the JobTracker console to have the tasktracker logs! Is there some way I could solve this?

    3. Is there a way to kill a Hive Job from the Toad console? The Stop SQL button in the SQL Editor stops the command from Toad but do not kill the job.



    Thanks,

    Mike
  • Mike at May 16, 2011 at 2:10 pm
    Thanks Peter!
    Mike

    --- On Mon, 5/16/11, Peter Hall wrote:

    From: Peter Hall <Peter.Hall@quest.com>
    Subject: RE: Using Toad for Hive
    To: "user@hive.apache.org" <user@hive.apache.org>
    Date: Monday, May 16, 2011, 1:29 AM




    #yiv653055334 P {margin-top:0;margin-bottom:0;}

    Hi Mike,



    1. There's no official support for Ubuntu, however I don't know of any reason why the hive features wouldn't work. The data hub is only available as a Fedora rpm, but is not required for direct hive connections.



    2. The jobtracker uses reverse DNS to determine the addresses of the tasktrackers. If it is getting this wrong you can set slave.host.name on each tasktracker or change your dns setup so reverse dns gives you the host name you want to use.



    3. Not at the moment. I'll pass this one up the chain and see if we can add it in a future release.



    Cheers,

    Peter Hall

    Quest Software



    From: Mike [nano_kollam@yahoo.com]

    Sent: Sunday, 15 May 2011 2:12

    To: user@hive.apache.org

    Subject: Using Toad for Hive










    Hi,

    I started using Toad for Hive,looks cool! I installed the plugins in Eclipse in windows and it started working fine.
    I have a few questions:

    1. Does it work with Ubuntu-Linux?
    2. Job Tracker works okay, but when I look into some tasktracker logs, the url gets changed as hostnames of the tasktrackers instead of IPs and so I have to manually edit the URL in the JobTracker console to have the tasktracker
    logs! Is there some way I could solve this?
    3. Is there a way to kill a Hive Job from the Toad console? The Stop SQL button in the SQL Editor stops the command from Toad but do not kill the job.

    Thanks,
    Mike
  • Mike at May 16, 2011 at 3:54 pm
    Hi all,

    I am trying to use PostgreSQL as the local metastore in Hive. I have the  postgresql-8.4-701.jdbc4.jar in the path. I use hive-0.7 in my system.

    I have the following properties added to hive-site.xml.


    <property>

    <name>hive.metastore.local</name>

    <value>true</value>

    </property>

    <property>

    <name>javax.jdo.option.ConnectionURL</name>

    <value>jdbc:postgresql://localhost:5432/hivedb</value>

    </property>

    <property>

    <name>javax.jdo.option.ConnectionDriverName</name>

    <value>org.postgresql.Driver</value>

    </property>

    <property>

    <name>javax.jdo.option.ConnectionUserName</name>

    <value>******</value>

    </property>

    <property>

    <name>javax.jdo.option.ConnectionPassword</name>

    <value>******</value>

    </property>

    I created a table using Hive Cli, and it went good. But when I try to

    DROP the table the hive cli hangs and is not responding.

    I found a similar thread here : http://search-hadoop.com/m/VYmqW1ejsiA&subj=Re+Can+t+drop+table , but I didnot quite get the solution. Can anyone help me on this?

    Thanks,

    Mike
  • Jov at May 17, 2011 at 6:03 am
    this is caused by HIVE auto create some meta tables when found them has not
    created.

    when you create table without partition or index,it will not touch the
    partition or index meta table,all goes well.but when you drop the table ,it
    will query the partition or index meta table,but find it has not created,so
    hive will create them,this ddl need some locks.this cause pgsql dead lock.

    there' three methods to solve it:
    1. connect to pg as super user, select * from pg_stat_activity,find the
    process id of backend which status is "<IDLE> in transaction",so the
    partition or index meta table will create sucess.then re run the drop talbe
    stmt.

    2.run some non-ddl stmt in hive:
    for example,after create table t,run:

    show partitions t;
    show index on t;

    these stmt will create the meta table but not cause dead lock.

    3.see:

    http://www.datanucleus.org/products/accessplatform_2_0/rdbms/schematool.html

    and use hive jdo file in src source_root/metastore/src/model/package.jdo as
    input file,to init the metastore db in pg.


    good luck!

    2011/5/16 Mike <nano_kollam@yahoo.com>
    Hi all,

    I am trying to use PostgreSQL as the local metastore in Hive. I have the
    postgresql-8.4-701.jdbc4.jar in the path. I use hive-0.7 in my system.


    I have the following properties added to hive-site.xml.


    <property>
    <name>hive.metastore.local</name>
    <value>true</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionURL</name>
    <value>jdbc:postgresql://localhost:5432/hivedb</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionDriverName</name>
    <value>org.postgresql.Driver</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionUserName</name>
    <value>******</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionPassword</name>
    <value>******</value>
    </property>

    I created a table using Hive Cli, and it went good. But when I try to
    DROP the table the hive cli hangs and is not responding.


    I found a similar thread here :
    http://search-hadoop.com/m/VYmqW1ejsiA&subj=Re+Can+t+drop+table , but I
    didnot quite get the solution.

    Can anyone help me on this?


    Thanks,
    Mike
  • 김영우 at May 17, 2011 at 6:23 am
    Hi Mike,

    It seems like there are some problems on creating metastore schema. So,
    create a schema for hive metastore first and then try it again.
    There is a DDL script for Pg,
    http://svn.apache.org/repos/asf/hive/trunk/metastore/scripts/upgrade/postgres/hive-schema-0.7.0.postgres.sql


    - Youngwoo


    2011/5/17 Mike <nano_kollam@yahoo.com>
    Hi all,

    I am trying to use PostgreSQL as the local metastore in Hive. I have the
    postgresql-8.4-701.jdbc4.jar in the path. I use hive-0.7 in my system.


    I have the following properties added to hive-site.xml.


    <property>
    <name>hive.metastore.local</name>
    <value>true</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionURL</name>
    <value>jdbc:postgresql://localhost:5432/hivedb</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionDriverName</name>
    <value>org.postgresql.Driver</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionUserName</name>
    <value>******</value>
    </property>
    <property>
    <name>javax.jdo.option.ConnectionPassword</name>
    <value>******</value>
    </property>

    I created a table using Hive Cli, and it went good. But when I try to
    DROP the table the hive cli hangs and is not responding.


    I found a similar thread here :
    http://search-hadoop.com/m/VYmqW1ejsiA&subj=Re+Can+t+drop+table , but I
    didnot quite get the solution.

    Can anyone help me on this?


    Thanks,
    Mike

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupuser @
categorieshive, hadoop
postedMay 14, '11 at 4:13p
activeMay 17, '11 at 6:23a
posts6
users4
websitehive.apache.org

People

Translate

site design / logo © 2022 Grokbase