To do a full establish on RBCV

find out devices logical name (symdg show) and then you can:

symmir -g Groupname establish -full DEV001 BCV ld BCV001

are you doing it from the R1 side ?

Thanks for the reply. We have 4 device groups and using srdf/a replicating data from production to DR site. there bcv's associated to the r2's in the remote site. the device groups has aleast 6 servers to make them consistent depending on applications. currently we are doing testing on those dr servers and one of the server sql won't come online. dba's came back saying there some corrupt data sets and I thought of doing a full resync of data instead of doing the incremental from r2's for that just one drive or rbcv device. all the rbcv's are part of a device group. so instead of performing on the whole group i was wondering if there is a way to achive it for one single device. I tried using ECC but I didn't not see an option for a full establish. If there anything through symcli/symmir would also be helpful. let me know for further questions.

The short answer is yes. The longer answer depends on what method you use to control SRDF/BCV operations in your environment.

If you use ECC you can control operations on individual devices within the GUI. If you use the SYMCLI command line you can create a DG with just the devices in question you want to manipulate and do the work through that. This is what I have done the most frequently and am comfortable/familiar doing.

If you are using that symrecover agent for SRDF then you have some extra steps.

So ... can you describe your environment a little more and what you are comfortable using and we can get you a better answer on how to establish/split just that single device.

What version of ECC are you using? In ECC 6.0 (what I am using - can not remember if 5.2 was the same) there is not an explicit "full" option but rather an "incremental" option which is checked by default. If you uncheck the "incremental" option you will get a full establish or restore.

Are you testing on the RBCV devices?

The alternate way is to make a DG with just the devices you want to establish and then you can use symmir -g DGNAME establish -full to run a full establish assuming you are testing on your RBCV and not your R2 devices.

I am sure that is what it will do. I'm less sure doing a full establish vs. an incremental will be any better than just doing the incremental restore. In theory an incremental will get you a point in time copy of the R2 - but you have to be careful that you are including EVERYTHING from the same time or the database will be appear to be corrupted.

Is the database completely contained on that device and if not is SRDF still active while you are doing this test?

What type of OS are you running on and what kind of database?

Recovery with SRDF/A should be very simple unless something is not setup correctly or perhaps someone is attempting to do something they should not with the given application.

srdf/a is active to R2's but the rbcv's are in a split state. there are 4 drives total to the particular server and i am planing to resync all the drives.

I think the command line should work , though not sure about the gui. but will give it a try.

sure ..so if you are running this agains the array that contains your R2s and BCVs and you are running from the host with SE that's connected on the R2 side ...that command should work.

i did the same thing on GBCV devices due to a failure of a hyper where those BCVs are asscoaited. But considering the DR recovery requierment i am not sure if you will be recover the DB(give a try though and do the cons split)

symmir -sid xxxx -f fix_bcvs_3-13-8 establish -full
symmir -sid xx -f fix_bcvs_3-13-8 split

[/var/tmp/emcscripts] head fix_bcvs_3-13-8
98a 939
9a4 c53