Steve Jones - Editor (1/12/2009)
Interesting you've never used it. I use this as a first glance when I get a complaint about a new server. It's a thumb-sketch to see if this is grossly out of whack.
Once I'm familiar with a server, I'd probably never look at it again.
Basically, I find that, if there's something slow about the server, I'm better off starting with either the specific complaint or with a trace that includes run-time.
If, for example, users are complaining that one of their pages is slowing down, I'd start with the queries on that page. If the whole system is slow, I'd start with a trace to find out what's eating all the resources.
Starting with "is BCHR low" just adds a step to either of those.
I mean, what do you do if a page is too slow, and you check and BCHR is high? What do you do if a page is slow and BCHR is low? In either case, you start looking at execution of the queries on the page. So why have the extra step in there.
What do you do if a server is slow and BCHR is low? What do you do if a server is slow and BCHR is high? You run a trace and find out which queries are slowing it all down, and handle them. Same handling in both cases. Why not skip the BCHR step?
Basically, I've not found a situation where it would help me narrow down my decision on what to do next. Maybe I'm missing something, but it hasn't come up for me yet.
How does it help you know what to do with a server? I may just be missing something.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon