VCF Multi AZ on VxBlock 1000 requires four (4) DR1s to be created.
- Two DR1s (datastores) for the Workload domain
- Two DR1s (datastores) for the Edge domain
Creating the DR1s
- Create volumes on both the AZ1 and AZ2 arrays as needed. For a DR1 at least one volume needs to be created in both AZ1 and AZ2.
i Metro node should be zoned in, and a Host, Storage Groups, Port Groups, and Masking Views have already been configured.
ii Under Storage select Volumes and click Create. Follow the steps in the wizard to create one or more volumes.
i Metro node should be zoned in, and a Host have already been configured.
ii Under Storage select Volumes and click +CREATE. Follow the steps in the wizard to create one or more volumes.
- Browse to one of the metro node appliances and login.
- Once logged in, on the navigation bar, click Provision Storage.
- Next to Provision on select cluster-1.
- Next to View By select Storage Arrays.
- Select the checkbox next to the storage array on which the volume was created.
- Click the More button and click Rediscover Array. In the confirmation window that appears click YES.
- A results window appears. Click CLOSE.
- Repeat this process for cluster-2 by selecting cluster-2 from the dropdown next to Provision on.
- To create the DR1, open the dropdown next to Provision on and select Distributed Storage.
- Next to View By select Virtual Volumes.
- Click CREATE and select Provision from Storage Volumes.
- The Provision from Storage Volumes wizard opens. Select an existing Consistency Group to add this DR1 too or select to create a new Consistency Group. Click Next.
- On the Volumes Options page, select the source Cluster. Leave the other options as they are and enter a name for the DR1 volume. Click Next.
- On the Storage Volumes page, select the storage array from the source cluster. Once selected it will show the available storage volumes. Select the one that was created for the DR1 and click Next.
- Repeat the previous step for the target storage array, and click Next.
- On the Storage Views page, select which storage view that will be used at cluster-1 and which storage view that will be used at cluster-2. If one was not set up previously, select None and create a storage view after the wizard is finished and add the DR1 to the storage view afterwards. Click Next.
- On the Review page, review the data presented and correct anything if needed. Once ready, click Finish.
- On the Results page, click Close to close the wizard.
- Back on the Distributed Virtual Volumes page, click the refresh icon to update the list of virtual volumes and their status. Depending on the size of the virtual volume, the health of the virtual volume will show as Minor Failure and the Operational Status will show as degraded. The status of the clusters will also show a warning symbol. This is normal during the replication process and can take from a few minutes to hours, depending on the size of the DR1 and the network connection between the clusters.
Procedure to speed up the replication of the DR1
Warning: the following procedure could impact the performance of the storage arrays, the metro node appliances, and the network connection between the metro node clusters. Proceed with caution.
By default, the DR1 is replicated at a Transfer Rate of 128K. Depending on the load of the storage arrays, the metro node, and the network connection between the two metro node clusters, the Transfer Rate can be increased up to a max of 128M by performing the following procedure.
- SSH into one of the metro node appliances.
- Enter the command:
- To list the distributed devices, enter the command:
ll (two lowercase ‘L’)
- Make note of the distributed volume you just created and enter the command:
- To list the status of the newly created DR1, enter the command:
ll (two lowercase ‘L’)
- A list of attributes for the DR1 is displayed. Notice the rebuild-eta and the transfer-size.
- To change the transfer size, enter the command:
set transfer-size <transfer-size>
example: set transfer-size 128M
- No response is provided to that command so run the command:
ll (two lowercase ‘L’)
to see the change and the progress of the synchronization of the DR1.
- When the DR1 has completely synchronized, run the following command to reset the transfer size:
set transfer-size 128K
Note: The recommendation is to slowly increase the transfer size from 128K to for example 2M, 10M, and so on, until you notice problems on the arrays, metro node appliances, or the network. Then change the transfer-size back down until everything is running normally again.