Re: Bug 98440 - oncheck -ci

In article <714khj$dfr$1@news.xmission.com>, Chuck Renaud <crenaud@informix.com> writes > >> >Try dropping and recreating the indices. >> >> On two tables with 14,000,000+ rows? The customer likes to be able to use >> their system from time to time! > >If the indices are corrupt, I don't imagine the customer is using >them too well anyway. True. But it would be jolly nice if we could use oncheck to point us a *which* indexes need attention - instead of having to drop and recreate *all* indexes. In fact this problem revealed something deeper which that 'UPDATE STATISTICS' seemed to be the cause of the corruption. I believe they are looking for better evidence to send to Tech Support in the UK. > >The problem seems to be reduced, but not totally eliminated, in >7.31/9.2, so an upgrade may help when those releases go GA. Rather a long way away at the moment, so it seems. <snip> > >For the latter, my personal experience is that 7.3 is much cleaner >than 7.2, though the HP-UX port is rather sloppy from a >documentation perspective. 7.30 *did* solve a major problem at another site which was a flaky network causing communication with the terminals to be lost and user threads to then bring the server down. So it's not all bad! This customer has a slightly earlier version of 7.30, UC3. I believe the one with the index problems has UC5. > -- Surfer! http://www.nevis-vieww.demon.co.uk http://www.nevis-vieww.demon.co.uk/flash Hopeful anti-spam: alter double 'w' to single 'w' to view site & send Email.