[Gluster-users] Stripe mode question

[eagleeyes]

Thanks a lot 
2009-04-16 



eagleeyes 



发件人： Raghavendra G 
发送时间： 2009-04-15  14:51:05 
收件人： eagleeyes 
抄送： gluster-users 
主题： Re: [Gluster-users] Stripe mode question 
 



2009/4/15 Raghavendra G <raghavendra.hg at gmail.com>

Hi, 

The actual total storage used across all the children of stripe will be the same as file size. Hence there is no extra storage

I was mentioning about the size got by du -s <file>
 
consumed, though ls -lh <file> on each backend reports size approximately equal to file size (which will make total storage to be equal to n * file-size, n being number of children of stripe).

sorry for the confusion.

regards,


2009/4/15 Raghavendra G <raghavendra.hg at gmail.com>


on the backend, striped files have holes in those regions data is not present. Hence size of the files on all nodes is the same. Advantage of stripe is the speed gained during reads and storage is not saved by striping.


On Mon, Apr 13, 2009 at 11:22 AM, eagleeyes <eagleeyes at 126.com> wrote:


Hello：
     i have a question about stripe mode ,when i use stripe like this 
        
volume bricks
  type cluster/stripe
  option block-size 1MB
  subvolumes client1 client2 client3
end-volume

    When i copy a file which 329M ,the file on  client1 client2 client3 is the same size ,but not three fragments ,the stripe mode  mean what ? 
    I want the file break into pieces which 1M size  

2009-04-13 



eagleeyes 


_______________________________________________
Gluster-users mailing list
Gluster-users at gluster.org
http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users





-- 
Raghavendra G





-- 
Raghavendra G





-- 
Raghavendra G
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20090416/88d60ed2/attachment.html>