BI Threshold not stalling database

[dlangschied]

I have a BI Threshold of 8.5 Gb on my OE 10.1C database of 28 Gb. Last week the BI grew to 9.0Gb and yet my database did not react by going into a stall. I am curious as to why this did not function. Any dieas?

 

[TomBascom]

I'd start by double checking the number of zeros specified in -bithold. The units are MB so for 8.5GB it should be about 8700.

It's also rounded to "the nearest cluster boundary" so if your bi cluster size is large that might also have an impact.

Lastly there should be something in your the startup section of you .lg file telling you what value was actually used.

BTW, -bithold does not stall the db. It causes an emergency shutdown unless you are also using -bistall.

 

[dlangschied]

I am using -bistall as well as the -bithold of 8500. I did, however, set this from proquiet. It does show in the log file that the threshold was updated, but I am not seeing it in the activity log (8648Mb) and certainly not on the OS where the bi extent file size totals roughly 9000.

[FONT=r_ansi][FONT=r_ansi][2011/04/16@22:24:56.805-0400] P-3400 T-0 I DBUTIL 18: (9242) BI File
Threshold size has been changed from 8.3 GBytes, to 8.5 GBytes.
[/FONT][/FONT]

 

[TomBascom]

I think it is probably just rounding. Your total bi extent size is close enough to your threshold and the exact numbers fuzzy enough that I suspect it only *almost* hit the threshold.

Plus bi extent size <> "bi in use". It is usually close to the same thing when the bi is growing but it isn't exactly the same thing.

What is your bi cluster size?