You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hadoop.apache.org by Aman Singh <bi...@gmail.com> on 2021/08/13 04:11:46 UTC

Re: [Urgent] - CDH upgrade v5.4.5 to v5.8 - Cluster to different Cluster upgrade

There are no issues moving within the same major versions.

In Hbase the challenge is only if we move from Hbase 1.x to 2.x. There we
have to take care of MOB, coprocessors, Prefix etc.

Anyway, moving/copying data across clusters does not cause issues. We only
need to worry about in-place upgrades.

When you say distcp, for Hbase do snapshot export.

On Thu, Dec 8, 2016 at 9:41 PM Viswanathan J <ja...@gmail.com>
wrote:

> Hi Guys,
>
> Currently we are running CDH version of 5.4.5 and planning to upgrade CDH
> v5.8 in different cluster. We are planning to migrate data to the new
> cluster through distcp after cluster upgrade.
>
> *CDH (v5.4.5) :*
>
>    - Hadoop v2.6.0
>    - HBase v1.0.0
>
> *CDH (v5.8) :*
>
>
>    - Hadoop v2.6.0
>    - HBase v1.2.0
>
> Is the latest version will be backward compatible with the older version
> of HDFS/HBase data?
>
> Also we are storing the data based on pre-split. Will this upgrade cause
> any data incompatible because of different version and cluster/zk nodes.
>
> Please help on this.
> --
> Regards,
> Viswa.J
>

Re: [Urgent] - CDH upgrade v5.4.5 to v5.8 - Cluster to different Cluster upgrade

Posted by Aman Singh <bi...@gmail.com>.
strangely, i see the question posting data as 2016, but this landed in my
inbox a few minutes ago :)


On Fri, Aug 13, 2021 at 2:11 PM Aman Singh <bi...@gmail.com>
wrote:

> There are no issues moving within the same major versions.
>
> In Hbase the challenge is only if we move from Hbase 1.x to 2.x. There we
> have to take care of MOB, coprocessors, Prefix etc.
>
> Anyway, moving/copying data across clusters does not cause issues. We only
> need to worry about in-place upgrades.
>
> When you say distcp, for Hbase do snapshot export.
>
> On Thu, Dec 8, 2016 at 9:41 PM Viswanathan J <ja...@gmail.com>
> wrote:
>
>> Hi Guys,
>>
>> Currently we are running CDH version of 5.4.5 and planning to upgrade CDH
>> v5.8 in different cluster. We are planning to migrate data to the new
>> cluster through distcp after cluster upgrade.
>>
>> *CDH (v5.4.5) :*
>>
>>    - Hadoop v2.6.0
>>    - HBase v1.0.0
>>
>> *CDH (v5.8) :*
>>
>>
>>    - Hadoop v2.6.0
>>    - HBase v1.2.0
>>
>> Is the latest version will be backward compatible with the older version
>> of HDFS/HBase data?
>>
>> Also we are storing the data based on pre-split. Will this upgrade cause
>> any data incompatible because of different version and cluster/zk nodes.
>>
>> Please help on this.
>> --
>> Regards,
>> Viswa.J
>>
>

Re: [Urgent] - CDH upgrade v5.4.5 to v5.8 - Cluster to different Cluster upgrade

Posted by Aman Singh <bi...@gmail.com>.
strangely, i see the question posting data as 2016, but this landed in my
inbox a few minutes ago :)


On Fri, Aug 13, 2021 at 2:11 PM Aman Singh <bi...@gmail.com>
wrote:

> There are no issues moving within the same major versions.
>
> In Hbase the challenge is only if we move from Hbase 1.x to 2.x. There we
> have to take care of MOB, coprocessors, Prefix etc.
>
> Anyway, moving/copying data across clusters does not cause issues. We only
> need to worry about in-place upgrades.
>
> When you say distcp, for Hbase do snapshot export.
>
> On Thu, Dec 8, 2016 at 9:41 PM Viswanathan J <ja...@gmail.com>
> wrote:
>
>> Hi Guys,
>>
>> Currently we are running CDH version of 5.4.5 and planning to upgrade CDH
>> v5.8 in different cluster. We are planning to migrate data to the new
>> cluster through distcp after cluster upgrade.
>>
>> *CDH (v5.4.5) :*
>>
>>    - Hadoop v2.6.0
>>    - HBase v1.0.0
>>
>> *CDH (v5.8) :*
>>
>>
>>    - Hadoop v2.6.0
>>    - HBase v1.2.0
>>
>> Is the latest version will be backward compatible with the older version
>> of HDFS/HBase data?
>>
>> Also we are storing the data based on pre-split. Will this upgrade cause
>> any data incompatible because of different version and cluster/zk nodes.
>>
>> Please help on this.
>> --
>> Regards,
>> Viswa.J
>>
>