hive-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nitin Pawar <>
Subject Re: Regarding Backup of MySQL if we have that as Metadata store for HIVE
Date Wed, 02 Jan 2013 14:30:55 GMT
if you are using hcatalog for metadata store, then you can just setup a
daily cron to take up data bacup with mysqldump.

from what I understand is keep the metadata backup setup from hadoop
operations as you dont really need to complicate it.

Just write a mysqldump query and assuming that you really dont have a huge
metadata store, you can make it hourly or twice a day backup and store in
and then you can easily restore from the dump files.

Others may have better solutions but this is how I had done it sometime

On Wed, Jan 2, 2013 at 7:51 PM, Ramasubramanian Narayanan <> wrote:

> Hi,
> If we have MySQL as the metadata store for HIVE,
> 1) do we need to backup every day ?
> 2) Is there any automatic way in Hadoop to have replication copy for MySQL
> too?
> 3) Anyway to update the metadata if we loose the information in MySQL?
> regards,
> Rams

Nitin Pawar

View raw message