Well, the OS is going to do what it is going to do, unless there are
specific commands/flags to alter its behavior. Your control, in this
instance, is that you are hitting the cache rather than requesting file I/O
( however the OS carries that out. )
But, what does that have to do with reliability? That is, why does Eric
seem to think that memory load is more reliable than disk load, which was my
question/comment. Particularly with paging, which is effectively disk load.
Joe Sam
-----Original Message-----
From: Martin Gloeckle <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: Tuesday, May 25, 1999 10:06 PM
Subject: Re: FW: file cache
>of course, as soon as paging comes in, that is something you can't control.
>
>but besides that, by running your own cache, _you_ control what is cached
>and what not, and not the operating system.
>
>- martin
>
>
>> -----Original Message-----
>> From: Joe Sam Shirah [mailto:[EMAIL PROTECTED]]
>> Sent: Tuesday, May 25, 1999 5:48 PM
>> To: [EMAIL PROTECTED]
>> Subject: Re: FW: file cache
>>
>>
>> Hi,
>>
>> I agree with pretty much everything you said, although things will
>> obviously differ from environment to environment and even run to run
>> number of files, file size, app activity, number of apps,
>> paging, and so
>> on. ) I do wonder though, why you think caching would be
>> more reliable
>> particularly when paging is brought into the picture. )
>>
>>
>> Joe Sam
>>
___________________________________________________________________________
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff SERVLET-INTEREST".
Archives: http://archives.java.sun.com/archives/servlet-interest.html
Resources: http://java.sun.com/products/servlet/external-resources.html
LISTSERV Help: http://www.lsoft.com/manuals/user/user.html
