Thanks to all those who responded. The best answer I got was to use DiskSuite to mirror swap or root disks, and you can dump to a metadevice. -- Greg Gallagher Sr. UNIX Systems Administrator First Options Chicago .... (312) 362-3643 On Wed, Apr 03, 2002 at 11:01:00AM -0600, Greg Gallagher wrote: > I recently had a problem on a SunFire E3800 running Solaris 2.8, > where I lost an entire IO Boat (i.e. the PCI IO controller and all > fiber and QFE cards on that boat). > > I have Veritas encapsulated disks which are mirrored across > controllers: > > c1t0d0 - primary > c0t0d0 - mirror > > My swapdevice, which is currently a Veritas encapsulated volume, > cannot be used to dump to, according to Sun because if the system > panics or has a Volume Manager problem dumping to swap isn't going to > help me. Sun suggested having a dedicated swap device, if we wanted > to have Veritas Encapsulated mirrored swap. > > The problem that I see with this is what happened to me yesterday. > Dumpadm was configured like so: > > excalibur:/root# dumpadm > Dump content: kernel pages > Dump device: /dev/dsk/c0t0d0s1 (dedicated) > Savecore directory: /var/crash/excalibur > Savecore enabled: yes > > This slice is one side of the swap mirror. However, swap is > configured like this: > > excalibur:/root# swap -l > swapfile dev swaplo blocks free > /dev/vx/dsk/swapvol 229,8 16 16383600 16383600 > > If I tell dumpadm to use swap, it refuses: > > excalibur:/root# dumpadm -d swap > dumpadm: no swap devices could be configured as the dump device > > Now what happens if /dev/dsk/c0t0d0s1 is inaccessible? Here's what > I saw on the panic(): > > syncing file systems... > panic[cpu8]/thread=2a100091d40: panic sync timeout > dumping to /dev/dsk/c0t0d0s1, offset 65536 > 0% done: 0 pages dumped, compression ratio 0.00, dump failed: error 5 > > Sure, when the system comes back up I can reset the dump device to > be the other side of the swap mirror, c1t0d0s1... but I've lost the > data from the panic. > > So my REAL question is this: How can I configure everything so I > have a fault tolerant dump device? Is it possible? > > Thanks in advance, I'll summarize, > > -- > Greg Gallagher > Sr. UNIX Systems Administrator > First Options Chicago .... (312) 362-3643 > _______________________________________________ > sunmanagers mailing list > sunmanagers@sunmanagers.org > http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagersReceived on Thu Apr 4 11:05:22 2002
This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:42:39 EST