HPE Storage Users Group

Finally getting around to setting up our third 3Par 7000 series array, initial config went ok but ended up getting stuck building up the Remote Copy target.



All hosts are running 3.1.2 utilizing RCIP. The HP7400-1 and HP7200-1-DR have been up and running with an Async RCIP config for a few months, I'm trying to add the additional 7200 when I run into the error.

The only code difference I'm seeing is the original two nodes are on 3.1.2 MU2 while the new one is on MU3, but I didn't see anything in the docs requiring matching builds between arrays.

I've never tried to do this myself, but with 3.1.2 there are many more restrictions for remote copy compared to 3.1.3.

The remote copy user guide states that if using RCFC then there needs to be 4 controllers.

It also states that if the primary system is a 2 node system then it must be either an S, T or F class system, so it's probably just that, it's not liking the 7400. If you go to 3.1.3 I bet it'll be fine, as that user guide says the 7000 is supported.

So I'd check the remote copy user guides for both versions.

Kinda figured, I've sent in my request to the power at HP (3par-sps@hp.com) to get upgraded on all three nodes this morning....we'll see how many weeks it takes them to get around to it.

It should be pretty quick, if you agree to their terms of checking the compatibility yourself then they will arrange it to happen within 3 days, if not then you need to fill out that awful spreadsheet with all the versions of firmware and drivers etc and then it's 7 days.

Good luck and please let us know if it solves the error you are seeing

Dk

After a few delays with the 3TB drives all systems have been updated to 3.1.3 MU1. Unfortunately, I'm still seeing the same error message. Any other ideas?

Hmmm, disappointing.

I've also seen this error when the ports needed to be reset or initiated I think it was.

So I'd maybe try that?

Also probably worth opening a support call with hp.

Ended up opening a support ticket. Per HP support since the original RC topology was 1-1, there is no migration method. So the "support" recommendation is tear everything out and rebuild from scratch.

Scripted out the tear-down and rebuild to save on the clicking but it kicks off a full re-sync.

FWIW, if anyone else encounters this issue try manually creating the additional links via the CLI and ignore the Java-based client, you might be able to modify the config without needing to rebuild the config from scratch.

Are you replicating the same data to both remote locations? let's say, VV1 goes into System B and C?

if so, how did you manage to include the same VV into two different replication groups? i'm stucked trying to replicate one source volume to two targets over IP.
thanks!