Great SQL Server Debates: Buffer Cache Hit Ratio. Really, is it one?

added by ab5sr
7/16/2012 9:30:53 AM

3 Kicks, 136 Views

Critique on a recent paper by a SQL expert, MVP, and MCM in our industry.


4 comments

dpeterson
7/16/2012 9:31:29 AM
BCHR is yet another in a long line of metrics that *can* tell you when your databases are performing well, but don't tell the entire story themselves. It is always important to monitor performance at the application layer as well, rather than relying fully on server metrics.

timbolt8
7/16/2012 10:02:02 AM
Eh, I am biased as an app developer, more often then not I find network and database issues to be more pertinant then any type of application errors/performance issues. I find this to be more of an issue if the application was archtected improperly from the very beginning.

dpeterson
7/16/2012 10:04:10 AM
Well, I didn't mean to blame the application. I just meant that rather than watching only performance counters to monitor application health/performance, we should be watching the application and using the counters to help diagnose issues when they arise, instead of treating them as indicators of performance issues.

timbolt8
7/16/2012 10:06:30 AM
Oh that makes more sense. I agree. You mean they shouldn't be counted on as full reliance....