Hi,
What is the size of the volume you are mirroring & how many such volumes ? I am assuming you are using cluster as you are mentioning about service group. Panic will only happen if PanicSystemOnDGLoss attribute of diskgroup resource is set to 1.
refer below for more details
https://sort.symantec.com/public/documents/sfha/6.1/solaris/productguides/html/vcs_bundled_agents/ch02s02s05.htm
If attribute is set to 1, & if diskgorup gets disabled because of storage issues in backend, then you have chances to get a server panic.
I am thinking in direction of IOFencing as well but it won't cause a panic while you are mirroring disks.
My suggestion would be to let mirroring continue if you are sure you storage connectivity is OK & you don't have issues reported for Storage. If the size of volumes is big & there are multiple volumes to mirror, it is normal for vxconfigd to take long time to finish mirroring as it will go sequential.
G