| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
Home -> Community -> Usenet -> c.d.o.server -> Re: Read Hit Ratio Question- Suggestions Welcome.
Here is a good page to check out:
http://www.flash.net/~darcre/dbtuning.htm
tmgn wrote in message <37A9C079.83270025_at_excite.com>...
>Hi Tapan
>Sometimes the Consistent Gets parameter gets so High (when the Instance is
>up and running for months together) that it skews up the Result..It
>reflects the accumulated value rather than the current scenario..
>
>There was one more issue related to Consistent Gets that it becomes a
>negative number sometimes which by the way is base bug 251434, fixed in
>7.2.
>
>The Better way to determine the Buffer Cache hit Ratio is to run the
>UtlBstat/Estat reports during the Peak time and then determine the Read
>hit Ratio.
>
>Also note that in 7.3.4 and Oracle 8 ,Changes were made to the 'physical
>Reads' parameter which now includes direct block reads as well as reads to
>get data into the buffer cache. Hence the above formula only gives an
>lower bound for the hit ratio on these releases.
>
>
>Hope this Helps..
>
>-Thiru
>
>
>
>
>
>Tapan Trivedi wrote:
>
>> Hey Gurus,
>> I am doing this query to find out the Read Hit ratios on a Baseline
>> system (upon which further development will be done) .
>>
>> select 1-(sum(decode(name, 'physical reads', value,0))/
>> (sum(decode(name, 'db block gets', value,0)) +
>> (sum(decode(name, 'consistent gets', value,0))))) * 100
>> "Read Hit Ratio"
>> from v$sysstat;
>>
>> I am getting very weird results on about four systems on which I have
>> run the queries . I am getting NEGATIVE, yes, NEGATIVE results. What
>> does this mean ?
>>
>> I am getting values like
>> -.047
>> -18.99
>> -2.31
>> -.69 etc.
>>
>> What does this mean ?
>>
>> Tapan
>
Received on Thu Aug 05 1999 - 14:09:11 CDT
![]() |
![]() |