RE: doubt about 99% of the cases if the first byte of the file is cached.

Hi, Ingo

In our test case, we just build 10 files and each file is 5MB,10 files and
each file is 10MB, 10 files and each file is 50MB. The main memory of
TUX server is 256MB.

Why  we build such big files for testing, the reason is we want linux to
flush out those files when the cache limit is reached.  So we can update
caching status table in proxy.

We already test the feasibility of our project when we will use
flush_inode_pages()
in our program to flush a speicifc file.  But now we want to get a real
flush
by Linux itself.

Now the question is can we know the maximum cache size of a backend
server running Linux 2.4.18 if the main memory is 256MB?

BTW, we record the time for the request of a speicifc file in our test case.
We found if the total size of all those http request files is bigger than
100MB, then the
request time for a file again and again will be increased instead of
decreased.
So the performance of TUX is not improved because Linux take much time to
swap between cache and disk.

But if the total size of all those http request files is less than 100MB,
then the
request time for a file again and again will be decreased. So the
performance
of TUX is improved because Linux already cached all these files in cache.

Is this result reasonable?

Liang



-----Original Message-----
From: [email protected] [mailto:[email protected]]On Behalf Of Ingo Molnar
Sent: Tuesday, April 23, 2002 7:43 AM
To: Liang Yang
Cc: [email protected]
Subject: RE: doubt about 99% of the cases if the first byte of the file is
cached.



On Tue, 23 Apr 2002, Liang Yang wrote:

> But even we didn't enable noatime, LED should flash only for 1/10 second
> or shorter time since Linux just need to update the access time flag of
> this file. it is not a big job and it shouldn't take Linux a long time
> to update this filed. But we noticed the Hard Disk LED flash for almost
> 2 seconds each time. So maybe Linux is doing some large amount hard disk
> read/write operation instead of just updating file access time flag.

well, is RAM large enough for all files accessed to be cached in RAM?

> I guess for all the sizes of files, the access time flag should be the
> same. Is that correct?

yes, but each file has a different access flag, in different on-disk
positions. So if you are accessing 100 files, the atime update sequence
might create a longer LED flashing.

	Ingo