Buffer wait ratio
Tony Woods asked what the buffer wait ratio formula bufwaits/(pagreads+bufwrits)*100 means. Art Kagel said it is the Bufwaits Ratio, the percentage of accesses that must wait for a buffer or LRU queue access, and distinguished it from buffer turnover. Murat Yildiz questioned Art's turnover formula and Art agreed it was inverted: (pagreads+bufwrits)/BUFFERS per minute.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
Hello all, Calculating Buffer wait ratio or Buffer turnover ratio : ((bufwaits/(pagreads + bufwrits)) * 100) What does the above formulla indicate? Thanks a lot for all your replies. Tony _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com
Tony Woods wrote: > > Hello all, > > Calculating Buffer wait ratio or Buffer turnover ratio : > ((bufwaits/(pagreads + bufwrits)) * 100) That's the Bufwaits Ratio (BR), yes. Buffer turnover rate is different: (BUFFERS / (pagreads + bufwrits)) / [minutes since stats zero'd] > What does the above formulla indicate? the BR indicates the % of accesses that have to wait for a buffer to write into either to update or to read from disk or to get access to an LRU queue to select such a buffer. Art S. Kagel > Thanks a lot for all your replies. > > Tony > _________________________________________________________________ > Get your FREE download of MSN Explorer at http://explorer.msn.com
In article <3A8109CF.3FC696BA@erols.com>, "Art S. Kagel" <kagel@erols.com> wrote: > Tony Woods wrote: > > > > Hello all, > > > > Calculating Buffer wait ratio or Buffer turnover ratio : > > ((bufwaits/(pagreads + bufwrits)) * 100) > > That's the Bufwaits Ratio (BR), yes. Buffer turnover rate is different: > > (BUFFERS / (pagreads + bufwrits)) / [minutes since stats zero'd] Shouldn't this be ((pagreads + bufwrits)/BUFFERS ) / [minutes since stats zero'd]? Or am I wrong? > > > What does the above formulla indicate? > > the BR indicates the % of accesses that have to wait for a buffer to > write into either to update or to read from disk or to get access to an > LRU queue to select such a buffer. > > Art S. Kagel > > > Thanks a lot for all your replies. > > > > Tony > > _________________________________________________________________ > > Get your FREE download of MSN Explorer at http://explorer.msn.com > Sent via Deja.com http://www.deja.com/
Murat YILDIZ wrote: > > In article <3A8109CF.3FC696BA@erols.com>, > "Art S. Kagel" <kagel@erols.com> wrote: > > Tony Woods wrote: > > > > > > Hello all, > > > > > > Calculating Buffer wait ratio or Buffer turnover ratio : > > > ((bufwaits/(pagreads + bufwrits)) * 100) > > > > That's the Bufwaits Ratio (BR), yes. Buffer turnover rate is > different: > > > > (BUFFERS / (pagreads + bufwrits)) / [minutes since stats zero'd] > > Shouldn't this be ((pagreads + bufwrits)/BUFFERS ) / [minutes since > stats zero'd]? Or am I wrong? Yes, thanks. That's what I get for answering postings at 3AM! Art S. Kagel > > > What does the above formulla indicate? > > > > the BR indicates the % of accesses that have to wait for a buffer to > > write into either to update or to read from disk or to get access to > an > > LRU queue to select such a buffer. > > > > Art S. Kagel > > > > > Thanks a lot for all your replies. > > > > > > Tony > > > _________________________________________________________________ > > > Get your FREE download of MSN Explorer at http://explorer.msn.com > > > > Sent via Deja.com > http://www.deja.com/