SRDF/Metro WP Update

Just a quick post to let you know I completed the update of my SRDF/Metro whitepaper to reflect both changes in the 5977 2016 Q3 SR and the SRDF SRA 6.2 release. In short, here are the general changes:

  • Included new functionality for virtual witness (vWitness). Though the environment I present in the paper uses a physical witness, I explain how a vWitness could replace it instead (or even in addition).
  • Updated naming convention changes. In the initial release of SRDF/Metro we called the witness a “quorum”. Here is an example:

quorumClick to enlarge – use browser back button to return to post

Of course the witness is a quorum, but we decided to change the naming to witness for consistency:


Click to enlarge – use browser back button to return to post

There are other places the name changed, too, but no matter what version of Solutions Enabler/Unisphere for VMAX you are running, just remember quorum=witness.

  • Mention of PP/VE 6.1 in relation to vMSC uniform configurations.
  • Explanation of how to add SRDF/Metro pairs online.
  • Discussion of the SRDF SRA 6.2 – in particular that SRDF/Metro configurations with witness in VMware SRM is now supported as well as 3-site, cascaded or concurrent setups.

As always I try to keep these papers up-to-date (within reason) so you have the latest information available as you configure your environments.


15 thoughts on “SRDF/Metro WP Update

Add yours

  1. HI Drew

    I’m not sure if I’m in a correct post. But I Have onde doubt about Auto Link Recovery with SRDF/S and SRDF/METRO.

    When happens a failure in SRDF link between the arrays. The volumes became partioned. And when the links are restablished. The status of volumes change to “Suspend” needing a manual intervention in the option “establish” to became “Syncronous” or “ActiveActive” again. If I enable “Auto link recovery” at local site, the process of suspend to establish, will be automatic? What concerns should I have enabling this feature for a automatice replica after link failure or a outage of remote array?

    Many thanks again

      1. So with Metro, there’s no way to start the replica automatic after a outage ? Should the user made the manual intervention and establish de SRDF/METRO ?


      2. It must be done manually and frankly it is much safer that way. You’ll want to figure out why the link went down before re-establishing.

  2. Ok, thanks for clarify this. I used to work with RecoverPoint Asyncronous and the replica is established automatic. But with SRDF is different.

    In case of stay 1 day without link, after the link returns, Should I just “establish” a SRDF/METRO group, or would be a good practice, change the group do Adaptive copy, or something similar?

  3. Great, many thanks My concern was about to many thing to replicate and this cause some delay to hosts. but the SRDF links are 16GB so, I thins is enough.

    Last question, I Promisse. I’m doing some tests before migrate the production enviroment to VMAX and I’m in a test situation where I had a Outage and Array A R1 became R2 and vice-versa. What should be the correct procedure to change Array A iremain as R1 and Array B as R2 ? I Tryed a swap/restore commands but in all situations I got error messages

  4. Thanks again Drew

    I’m using a vwitness protection

    If I Understanded correctly, Should I Suspend the SRDF Group then change the protection from witness to bias an then re-establish the pair, changing R1/R2? Or Is there a “non-disruptive – R2” way to perform this operation?

    1. You have to use suspend with the -keep option when utilizing the witness to change the R1. The Best Practices Metro paper has the syntax.

  5. Great Tip Drew

    With that tip, keeping R2, I Got more availability and one step less :) in restablish operations

    One doubt about Powerpath in Solaris. I’m having a strange behavior. When I mask one new LUN to a Solaris host, with powerpath 6.0, the LUNs does’t come available to host, with commands like “powemt restore/check/confi” – I Only got sucess when I shut down and power on the FC interface. Do you know if there is as specific configuration, to perform this recognition automatic in Solaris? I’m using a Portgroup with 8 paths(4 per director) to present LUNs from VMAX to Solaris Oracle ASM I’m not sure if 8 paths is ok or to much for it

    Many thanks again

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Powered by

Up ↑

%d bloggers like this: