RE: [U2] UV 10.1.2 bug

2005-01-18 Thread Brian Leach
Sounds like the kind of thing that would happen if the file had a lot of overflow. Or a very badly designed index, or trigger. Any of those possible? Brian -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of [EMAIL PROTECTED] Sent: Monday, January

RE: [U2] UV 10.1.2 bug

2005-01-18 Thread karlp
quote who=Brian Leach Sounds like the kind of thing that would happen if the file had a lot of overflow. Or a very badly designed index, or trigger. Any of those possible? Brian We don't have any triggers on those files. There are no indexes on one of the files and the other one has 5

RE: [U2] UV 10.1.2 bug

2005-01-18 Thread karlp
Okay, tested the process against the file with no indexes. It ran in 19 minutes. it was taking hours before. I didn't save a copy of LIST.INDEX FILE ALL DETAIL but it did have some very skewed groups, as I watched it scroll by. But mostly it looked very even... I've never seen this behavior

Re: [U2] UV 10.1.2 bug

2005-01-18 Thread Clifton Oliver
I've seen this kind of thing when an index of a sparsely populated field was created without the NO.NULLS option. -- Regards, Clif ~~~ W. Clifton Oliver, CCP CLIFTON OLIVER ASSOCIATES Tel: +1 619 460 5678Web: www.oliver.com

RE: [U2] UV 10.1.2 bug

2005-01-18 Thread John Jenkins
: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: 18 January 2005 19:37 To: u2-users@listserver.u2ug.org Subject: RE: [U2] UV 10.1.2 bug Okay, tested the process against the file with no indexes. It ran in 19 minutes. it was taking hours before. I didn't save a copy

RE: [U2] UV 10.1.2 bug

2005-01-17 Thread Ed Clark
Sounds like the kind of thing that would happen if the file had a lot of overflow. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of [EMAIL PROTECTED] Sent: Monday, January 17, 2005 12:10 PM To: u2-users@listserver.u2ug.org Subject: [U2] UV 10.1.2 bug I'm

RE: [U2] UV 10.1.2 bug

2005-01-17 Thread karlp
quote who=Ed Clark Sounds like the kind of thing that would happen if the file had a lot of overflow. That's what I would have thought, too. However, I'm seeing this on files that have recently been resized to about 10% larger than current need for growth. AND they aren't that big. 30,000 items

RE: [U2] UV 10.1.2 bug

2005-01-17 Thread Ed Clark
Could they be hashing very badly? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of [EMAIL PROTECTED] Sent: Monday, January 17, 2005 2:30 PM To: u2-users@listserver.u2ug.org Subject: RE: [U2] UV 10.1.2 bug quote who=Ed Clark Sounds like the kind of thing

RE: [U2] UV 10.1.2 bug

2005-01-17 Thread John Jenkins
Definitely looks like overflow - what does an ANALYZE.FILE tell you? --- also what is the physical file structure (O.S level files and sizes). One other possibility - is there a horrendously large indexed entry (like 10's of K multivalues or similar) ? - let's take a look at LIST.INDEX ALL