[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Does fsfs revprop packing no longer allow usage of traditional backup software?

From: Ivan Zhakov <ivan_at_visualsvn.com>
Date: Fri, 1 Jul 2011 19:36:48 +0400

On Fri, Jul 1, 2011 at 19:04, Peter Samuelson <peter_at_p12n.org> wrote:
>
>> > Especially since the "common case", as you put it, is that the open()
>> > returns ENOENT, file not found.  _That_ case may or may not be
>> > expensive, but it's quite different from the usual consideration of
>> > whether stat() is expensive.  Not least because stat() is usually
>> > considered in context where a file _does_ exist and various information
>> > about the file gets looked up to populate the stat structure.
>
> [Ivan Zhakov]
>> Btw on Windows dry open() takes a lot more time than stat(). I don't why.
>
> Wait ... you're saying open() that returns 'file not found' is more
> expensive than a stat()?  It is harder for the OS to determine that a
> file does not exist and thus cannot be opened, than it is to gather
> various fields of information about a file that _does_ exist?
>
> I hate to pick on Windows for its design, because that's so cliche ...
> but that sounds like a pretty poor filesystem design.
I was wrong: we found this issue on ext2fs. See r1103749:
http://svn.apache.org/viewvc?view=revision&revision=1103749

I didn't tested difference in performance between stat() and open() on
Windows, but Stefan2 tested in on ext2fs.

-- 
Ivan Zhakov
Received on 2011-07-01 17:37:44 CEST

This is an archived mail posted to the Subversion Dev mailing list.