3PAR two storage remote

[Dave5]

One of my customers has a 3PAR connected to a Blade Enclosure (Production System @ Primary Site). Also they have a 3PAR at the Second site.

Would it be possible that if the 3PAR Storage Box @ Primary Site goes down, The 3PAR at the Second site would take up the load from the Production System. Is such a connectivity possible ?

Distance between Primary site and Second site is 3 KM

[Sheldon Smith · Accepted solution]

It's certainly possible.

• Both arrays would need to be licensed for HPE 3PAR Remote Copy.
• There are restrictions on the 3PAR OS version running on each.
• They might be able to link up via fiber channel (if there are unused ports), or you can use the 1Gb RC-1 (Remote Copy) ports on the controllers to use IP.

Depending on the hosts' operating system(s), there may be additional feature sets that can be added.

Note: While I work for Hewlett Packard Enterprise, all of my comments (whether noted or not), are my own and are not any official representation of the company.
----------
If my post was useful, click on my KUDOS! thumb below!

[Dave5]

Thanks Sheldon,

3par remote copyt is good. How would I then connect my other 3par to the server which is sitting 3kms away.

So we have one server already connected to the 3par and now connect aother 3par to the same server. like you said is this possible via FC or are there any other ways ? 

[Dave5]

Can the source and target storages be interconnected  and just the source storage to the server ?

[Dennis Handly]

>Can the source and target storages be interconnected

RC needs the two arrays connected.  See Sheldon's post for types of connections.

[Dave5]

If the Source array is down, will the target array be up and running ?

[Sheldon Smith]

Remote Copy needs two arrays. A host writes to the Primary volume and it is copied to the (duplicate) on the Secondary array.
Depending on round-trip latency times, it may be possible to do Synchronous RC, where each write is duplicated in real-time to the secondary array. If latencies are too long, the alternative is Asynchronous Periodic RC, where the changes are bundled up and sent to the secondary array's volumes on a periodic basis (hence the name).

Hosts can be connected over limited distance to the arrays; how depends on the arrays' ports. The arrays have FC ports built-in. They can have a pair of add-on cards added. The add-on cards can be FC or iSCSI. (The arrays may already have add-on cards.)
If FC, the hosts would need to be connected to FC switches which would in turn be connected to additional FC switches at the other site. Then using zoning, the hosts would see both storage arrays.
If the arrays have add-on iSCSI cards, the hosts could use iSCSI instead.

The point of Remote Copy is the source and target (primary and secondary) storage arrays are interconnected. If the primary array is down, a server would need to have access to the secondary array to continue processing.
Host is connected to primary, host is connected to secondary, primary replicates to secondary, host normally works with primary. Something happens to primary, replication can fail over, host can now work with secondary. When primary returns, the secondary and primary play "catch up", and eventually replication fails back to the original primary.

Contact Hewlett Packard Enterprise to discuss further with Sales and also Storage Consultants.