I have a application with a large number of small files, it would appear from the disk cache counters that the files are generally not getting cached for repeated reads.

I suspect that this is because the read load is distributed across so many files that very few of them are being accessed frequently enough or consistently enough that the OS believes they merit caching.

I have a plan to consolidate the contents of the files to bring down the number. Making the files too large will bring it's own problems. So I want to consolidate to the minimum point which would improve the caching performance.

In what way do the following factors affect Windows Server file caching behaviour: - The number of files accessed during a period - The size of the files - The number of times a specific file is accessed during a period - Whether a file has just been read or also written to

... and what factors have I missed?

link|improve this question
1  
What makes you think that the files are not being cached? What figures for cache hit rate does Perf Mon show? – Richard Sep 15 '10 at 10:44
I think that they aren't being cached because the cache size is climbing more slowly than data is being read in by the application and the cache hit rate is lower than the rate application indicates it is repeating file access at. The application is reading the whole file in each time, so partial reads do not account for the differences. – Bell Sep 15 '10 at 16:54
feedback

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
or
required, but never shown

Browse other questions tagged or ask your own question.