srdf cascade - recreating pair fails after 2nd hop is reinstated

Hello
we're testing srdf in a cascade locallly on our camput via fibre and then via GigE to a remote vmax.

All was going fine, everything in sync. we decided to test the replicated meta at the R21 stage by emoving the second pair of the hop, ie the pair / group corresponding to our asyn hop to the remote array. All fine. The R21 meta became R2 and was eventually mounted on our soalris host under zfs.

So, we 've failed over onto R2 - we decide to reinstigate replication to our remote site by reissueing exactly the same commands to create the group and make up the pair as the device numbers are of course the same. i.e reinstate the group and pairs relating to the second part of the hop.

symrdf addgrp -label fred -Gige  -sid xx -remote_sid xx -rdfg 8 -remote_rdfg 8 -dir 13g,13h,4g,4h -remote_dir 7g,7h,8g,8h

using same rdfg number and label as before .....all ok.

However, when attempting to recreate pairs

# symrdf createpair -file fred_R21_R2.txt -type R1 -sid xx -rdfg 8 -invalidate r2 -g fred

'-g': Incompatible option.

Device 2901 is already a member of group fred

Mmm. But it was a member of the group first time around ..... and it worked fine.

Appreciate others view on where we are going wrong here Thanks.