01:02:41 <jclulow> I wonder if we could do something like the DTrace buffer switch and relatively aggressively pull the kmem logging out to disk files
01:03:08 <jclulow> And you could then point mdb at the log file as well as the crash dump
03:08:38 <jbk> do you know from the stack what what code was freeing it?
10:21:37 <jclulow> vetal: I don't see, e.g., c39t5000CCA04D1A6BA9d0 that you mentioned earlier in that output
10:23:39 <jclulow> vetal: Can you run a few more commands, at any rate: "/usr/lib/scsi/sestopo", "iostat -En", "mpathadm list mpath-support", and "mpathadm list initiator-port" ?
10:24:47 <jclulow> And, I guess, "/usr/lib/fm/fmd/fmtopo -Vp" as well