HBase Master: Failed to become active master

[#1]

So ZK is working fine. Perfect.

Make sure HBase is down and clear everything:

hadoop fs -rm -r /hbase/*

echo "rmr /hbase" | zookeeper-client

Then try to start HBase.

JMS

[#2]

Hi JM, that worked!!

I am now able to startup HBase with all its servers!

Thanks for your assistance and persistance 🙂

Cheers,

TS

[#3]

You're very welcome!

Glad it now works for you. Enjoy your cluster!

Best,

JMS

[#4]

I'm new to Cloudera Cluster.  I read somewhere that you should not use "hadoop fs -rm -r /hbase/*" in a production enviroment.  Can you please give me detail what "hadoop fs -rm -r /hbase/*" does?  Is there any data lost or data remove to any of the services or database? 

Thanks,

-sr

[#5]

I have same doubt what happens when i delete "hdfs dfs  -rm -r /hbase/*" will lost data/namespaces in HBASE or HDFS?

[#6]

Hi Sandeep,

This command will remove ALL HBase data... Snapshots, tables, WALs, etc. So be very careful with it.

JMS

[#7]

so when the above command delete everything i will lost all HBASE tables/WAL's. Is the above steps recommened in Production environment?

[#8]

It depends what you want to achieve, right? If you goal is to repair your production cluster, then it might be relevant. If you goal is just to test and see  what it does, then what ever command it is, it should not be tested in production?

Basically, the command just does what I described. It deletes everything. It's you to decide if it's relevent for your production environment or not. If you goal is just to drop a table, or all tables, then the shell is better.

Are you facing any issue with your production environment?

[#9]

No...we don't have any issues so far with HBASE. I am just trying to understand the concept of it.

Thanks for your assistance 🙂

[#10]

Hello,

If the data was present for the same issue, then what is the best way of sollution ?

Can you please help me.

Thanks