Re: "Changing data structure forced command termination."

From: "Alnis Bajars" <alnisb@colorcorp.com.au> > >My client site has IDS Workgroups 7.30.UC9, and we're on HP-UX 11.0, and >the >server is an R390. > >The drama I am talking about happened when I was on holidays for 2 weeks, >so >I apologise if the details are a bit scant. > >The server "hung" on a couple of occasions with hardware related problems >(motherboard/memory), so the database server was "ungracefully" shut down, >on at least two occasions I can think of. > >An oncheck by an external consultant on Fri. 7/1/00 did not reveal any >problems viz data corruption. > >Two residual problems. >(1). OLTP performance is fine. But running certain reports brings the >database server literally to its knees. This was not happening before >these >episodes. >(2). This is more the issue. The error message in the title of this mail >happens when running "onstat -u", "onstat -k", "onstat -x", to name but a >few - these commands are partially executed before this error message >appears. "onstat -g seg" and "onstat -g glo" are okay. This happens >irrespective of whether shared memory or TCP/IP is used for the connection. > >I can't find anything in the obvious files (eg. online.log) to give me any >clues. > >My questions. >(1). Does anyone have any suggestions as a "first port of call" to try. Do you have more than 4 shared memory segments (type R and type V) >(2). Are there any files anyone would like to see? output from onstat -g seg. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com