Hey Guys!
Currently we have all our HS X Nodes, one Library, in one Data Center. Now the customer wants the backuped data also available on a secon site, where they want to build up a second HS X cluster, if DC 1 is down, all backup data is available in DC 2.
Solution 1, warm standby:
How can i push the protected data to the 2nd HS X Cluster besides running Aux copies? Is maybe a CV HS X Feature on the backend for that solution available?
Problems with Solution 1: The HSX Nodes are mostly idling, and the primary Cluster has to do all the workloads.
Solution 2, striped/cloned HS X Cluster:
The HS X Workloads are striped over DC 1 and DC 2, and all protected data has to be available on both sides. Benefit: We can distribute all Workloads over more HS X Nodes.
So does exist some kind of (Hedvig) Storage Data replication, to sync two HS X Libraries?
Or what solutions would you recommend as the “best” in this case?
Thank you!
Kr
Hello
Aux copy is going to be the best answer for you here. With the “DASH” feature being by default enabled it will only move each unique block once over the network and save a heap of network/write traffic.
This will work better than any replication tool if any existed. Which they do not.
Kind regards
Albert Williams
Hello
Aux copy is going to be the best answer for you here. With the “DASH” feature being by default enabled it will only move each unique block once over the network and save a heap of network/write traffic.
This will work better than any replication tool if any existed. Which they do not.
Kind regards
Albert Williams
Hi!
Thanks.
In this case, the 2nd HSX Cluster will only idle because the nodes only receive auxiliary copies, resulting in a significant waste of computing power.
Can we expand the current library, DDB, and all running tasks to the 2nd cluster, but with the feature that allows every task to run continuously without interruption if one Datacenter/HSX Cluster goes offline, allowing the 2nd Cluster to take over? In a single HSX Cluster, you can lose Nodes. So, is it also possible to lose a full cluster if you have two? Or, in this case, do I need a storage solution (for example, from Supermicro) that can perform dual writes?
Hello
If you goal is to have a HSX Cluster on standby in the case the active goes offline you will always have a heap of wasted compute.
If you have your AUX copy up-to-date, in the case of a DR you will just simply swap the primary copy and the secondary copy so that new backup jobs will be targeting the secondary cluster instead and backup will carry on like nothing changed.
If you want to have this in writing and confidence that your requirements have been taken into consideration i would recommend a PS architectural review to validate what you are wanting to achieve can be achieved.
Kind regards
Albert Williams
Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.