Re: Recovery from a disk failure on a mirrored drive

I'm with David on this one... Effectively, if you are using hardware mirroring ( not Informix ) then Informix shouldn't even know that anything has happened if the main chunk in the mirrored pair goes down. I've configured a server using hot swappable disks, mirrored at the OS level, pulled one of the disks out and hey presto - Informix carries on working ... How did you set up the logical volume ? Usually, within smit, you create the logical volume with a number of physical disks. Let me know the outcome ... Merry Xmas Sean ssoman@omm.com wrote in message <75ea03$dbi$1@nnrp1.dejanews.com>... >Hi again, Maybe this should be a question for the AIX user group rather than >the informix, but here it is. I am running INFORMIX 7.23 UC1 on an AIX ( >4.1.4 ) box. All the logical volumes ( which correspond to chunks ) are >mirrored at the physical level, ie.,for each disk partition used in a logical >volume, there is a copy on another disk. The problem is, that each time a >disk fails, INFORMIX detects an error and fails to move to the mirrored >physical disk. Which basically means that INFORMIX aborts itself, or marks a >chunk as down. Now, question is: Whose responsibility is it to switchover to >the mirrored disks, the controllers, the OS or INFORMIX ? > >As usual, thanks for the response(s) > >Sujata > >-----------== Posted via Deja News, The Discussion Network ==---------- >http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own