mccurdy@ucselx.uucp (mccurdy m) (05/17/89)
Hello, Several times, we have had a situation where IIDBDB becomes inconsistent. I don't have a clue. We are running INGRES 6.1 on VMS V4.7. I am able to "recover" the master database, but I know this problem will arise again unless I can get some ideas on why this is happening. I have heard that some bad things may happen when a user CTRL/Y's out of something. Anyone? Thanks ... Mike McCurdy |"Grandma died at 58 (619)284-7781 at home | Her best friend cried 3 hours straight (619)594-4653 at work | She had eyes like a little girl's mccurdy!ucselx!sdsu!ucsd | And by the way, who got the pearls?"
cs_bob@gsbacd.uchicago.edu (05/17/89)
> > Hello, > > Several times, we have had a situation where IIDBDB becomes > inconsistent. I don't have a clue. We are running INGRES > 6.1 on VMS V4.7. I am able to "recover" the master database, > but I know this problem will arise again unless I can get > some ideas on why this is happening. I have heard that some > bad things may happen when a user CTRL/Y's out of something. > Anyone? Thanks ... I definitely do NOT think that this problem arises when someone exits a program unexpectedly, via CONTROL-Y, or in any other way. The Ingres server is designed to address this very sort of problem, which was a major headache in version 5 and earlier releases. In 6+ months of version 6 usage, we have never seen the IIDBDB go inconsistent. We have seen other databases go inconsistant twice during that time, and in both instances the inconsistancy was in some way directly related to the mysterious death of a database server in a two-processor, one server per processor, cluster. The first obvious thing to check is the file II_CONFIG:ERRLOG.LOG . This may give you a clue to the errors that occurred preceding the inconsistancy. Otherwise, I think you'll need the help of RTI tech support. Bob Kohout