[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] Re: cow implementation


  • To: xen-users@xxxxxxxxxxxxxxxxxxx
  • From: Molle Bestefich <molle.bestefich@xxxxxxxxx>
  • Date: Wed, 25 Jan 2006 13:10:32 +0100
  • Delivery-date: Wed, 25 Jan 2006 12:19:27 +0000
  • Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=lG8ULKQ80oAsvuEfeH3jdjs+fRi9Ffm1Y733OaIhIHvG0+mvz/JrzBtGAcckW7nkm3AmCrHu4QNsARyaSTd0NPwyaE+EqTf+OgkVcMV8uIO1jqP3SCosY13bnMajCJK2qiBDnx+2+Wft3i9YStWDNNIOv5YS7uuHjnmvR57i/6Y=
  • List-id: Xen user discussion <xen-users.lists.xensource.com>

Steven Anderson wrote:
>  I would really like to see someone with experience on this start a wiki. I
> think a  lot of people could benifit from it.

That would be nice.
I'd be happy to contribute my experiences when I get more time.
(Which will probably be a couple of months from now, but anyway.)

Personally, I'd like to see some more information on filesystems that
zero blocks when they're freed and COW implementations that nukes
blocks that gets zeroed from the write target :-).

>  I also wonder what kind of performance and overhead a COW setup would have?

I've seen suggestions that unionfs adds about 15% overhead in a test setup.
It's of course dependant on your setup..
The more filesystems you add on top of each other, the slower it will
get, I imagine.

OTOH, dentrys are usually cached, so it might not be that big in the
common case.

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.