Help copying the Database to Another Cluster

[chenshaulian]

Hi all,

We have two 10 nodes Vertica environments (version 7.1.1-10). One is production and the other one is for DR.
I want to duplicate the production environment to the DR site.
The DR site was our DEV site, in time the environment became not updated any more (data and structure).

I know there are several ways, what do you think about those two ways:
1. CopyCluser, using vbr.py utility.
2. copy_vertica_database.sh script.

Can anyone recommend on one way than another?
What are the different between those two ways? What is better?

We have differences between the names of the nodes in the two environments (prod environment: node-a1, node-a2, node-a3…. And in the DR environment: node-b1, node-b2, node-b3…), I know it's a problem for the CopyCluster, right?
And I don't know if it is a problem, but we have dbadmin user in booth environment but with a different password, it can be a problem?

Thanks for your help
Chen

[Chicago_vertica]

1. they are similar. Just different scripting.
2. I use: /opt/vertica/bin/vbr.py --task copycluster --config-file /opt/vertica/config/copycluster.ini
3. It is not the node name (server name). It is the Vertica node name. Run this in both target and source: select node_name,storage_path, storage_usage from disk_storage;
They need to be the same.
4. I had a lot of trouble to get mine started. The key things are: 1- Passwdless SSH--source and target node; 2- Port 50000 should not be blocked; 3- Remove all the files at the target /tmp/vertica.

Good luck.

Yongzhi

[chenshaulian]

Thanks Yongzhi,

This is my problem, I have different Vertica node name between the environments.
can I use the copycluster task? or I need to search for other solution?

Thanks
Chen

[Chicago_vertica]

According to the document, copy cluster will not work.

Can you post the results from the source and target: select node_name,storage_path, storage_usage from disk_storage; ? I think there is a good chance for them to be the same ( it is DR cluster, why wouldn't you want them to be the same?). You will be far better off to reinstall Vertica if not the same.

I don't know any better ways to copy a cluster over. I think from the storage side you may try to snap the volume and mount it to the DR cluster. But it is not supported.

Your best option is Copy-cluster. It is fast and reliable. You can schedule to shutdown the database (before copy) and restart it (after copy).

Yongzhi

[NorbertKrupa]

Take a look at my blog post on the various methods to move data around and their limitations.

The recommended approach with a primary/DR configuration is to do dual ingestion. If you're looking to do a delayed restore from backup, you may want to consider upgrading to 7.2 as the requirements to have identical cluster topology or names have been removed.