On Monday 16 February 2004 11:13 pm, Greg 'groggy' Lehey wrote: > On Monday, 16 February 2004 at 12:09:35 +0100, Stijn Hoop wrote: > > On Fri, Feb 13, 2004 at 11:55:41PM +0100, Lukas Ertl wrote: > >> I have a very very far fetched idea. Could you please try this: > >> > >> *) cd /usr/src/sbin/vinum > >> *) vi commands.c > >> *) locate the function vinum_read(), the first line says "int error;" > >> *) initialize this var to zero, i.e. "int error = 0;" > >> *) recompile vinum and tell if that changed anything > > > > OK, more info now that I am at the console of the machine again. > > > > During the weekend I discovered that I had saved a copy of my > > previously working kernel somewhere in my homedirectory, so this > > morning I first compiled a Feb 6 version of /sbin/vinum, and then > > booted off the old kernel + vinum module, and executed the old > > /sbin/vinum. No error message, but the only things listed in my > > configuration were the drives (so no volumes/plexes/subdisks). > > Sorry, yes, I was going to reply to the previous message. The > dumpconfig output suggest that everything has been overwritten. I > fear this is the result of something like doing a saveconfig when > there was nothing to save. > > > My current theory is that I somehow goofed up during the kernel > > rebuild-a-thon last friday, and that I somewhere loaded the old > > vinum module alongside the new kernel; and this bit me by > > overwriting my configuration. > > A possibility, but unlikely. I'd honestly guess finger trouble. > > > Although the data that was on the volume is not crucial, it would be > > nice if there was a way to recover it; so if you don't want to > > investigate further, can you tell me whether a vinum create would > > restore my volumes, > > Yes, it should. The data should be unchanged, so the file system > should still be OK. > > > or is there simply no way that I can do so without obliterating the > > data? If the latter, I'll just move on and recreate the plex using > > ccd(4) (seeing as following vinum development is not one of my goals > > right now). > > It occurs to me that we could easily prevent this kind of problem in > the future by saving the Vinum configuration at boot time to > /var/run/vinum.conf.boot or some such. It would obviously make sense > to keep the previous configuration as well. This should be a pretty > trivial modification to the startup scripts. Haven't followed this too closely, but what if /var is a vinum volume? /dev/vinum ? > > Greg > -- > See complete headers for address and phone numbers. -- ian j hart http://ars.userfriendly.org/cartoons/?id=20031016Received on Wed Feb 18 2004 - 13:31:40 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:43 UTC