Re: 5.0 bugs (Was: tbcheck core dumping)

In article <BuspuC.IF1@portal.hq.videocart.com> kdenning@portal.hq.videocart.com (Karl Denninger) writes: References: <1992Sep15.010353.7889@anasaz> <HOLDER.92Sep18093635@gourde.positron.med.ge.com> <1992Sep18.212406.9261@adaclabs.com> Date: Fri, 18 Sep 1992 22:44:36 GMT Lines: 19 In article <1992Sep18.212406.9261@adaclabs.com> mlee@adaclabs.com (Mike Lee ) writes: >In article <HOLDER.92Sep18093635@gourde.positron.med.ge.com> holder@gourde.positron.med.ge.com (Kacy Holder) writes: >> >>data. 5.0 has a few *much* more serious problems we would >>rather they fix first... >> >Our company is planning on upgrading from 4.0 On-Line (SUN SPARC) to >5.0 On-Line. Can you mail or post a description of your known 5.0 bugs? > >Thanks in advance > Please POST. We run a large production database on Online 5.0 and have seen some strange stuff with it. PLEASE let us know what we're in for here! Okay, I'll post. We currently have around 6 open cases with Informix support most of which we think are related to corruption of pointers to BLOBs in the database. We make extensive use of BLOBs both in table space and blob space so most of you will probably not run into this. The mildest symptom is tools like tbstat and tbcheck core dumping. The worst is the tbinit daemon core dumping as it comes up. Informix just sent us a patch for that problem (tbinit core dump). We have also seen a variety of error messages when we run tbcheck -cD. Don't panic. On the whole we've found 5.0 to be a very stable release and when we have run into problems Regency Support has been very helpful. Our application stretches the Informix DBMS in some unusual ways so the majority of you will never run into these bugs. We have confidence that Informix is working hard to resolve the problems as quickly as possible. But hey! This is *software* folks. There will always be problems. What you look for is is a good company to back you up when you do run into those problems. I remain a happy Informix Customer. -- Katharine C. Holder Computer People Unlimited, Inc. email: holderk@positron.med.ge.com "These opinions are my own and not necessarily those of the client for whom I am working or Computer People Unlimited, Inc."