[Gluster-users] cluster/afr not mirroring 100%?

Keith Freedman freedman at FreeFormIT.com
Sat Oct 11 09:53:25 UTC 2008


the block sizes would more depend on the file system you're running, 
not necessarily the underlying block device (lvm/raid/etc.)

it's possible fc and centos have different defaults for block sizes.
Also, depending on the file system, some will default to different 
block sizes depending on the size of the file system, so if these 
differ that could account for the difference in reporting.

I don't think it would be du or find that's behaving 
differently.   I'm using AFR across fc9 32 & fc9 64 bit.  with 
different device sizes on each end and different block sizes, so I 
can't rely on du to be exact.

glad it's all working, in any case :)  I was happy with glusters 
performance through my recent disaster too.

if you've not already, make sure you list yourself in their "whose 
using gluster" wiki, it'll help them get funding so they can continue 
their great work.



At 02:37 AM 10/11/2008, Jason Haar wrote:
>Krishna Srinivas wrote:
> > Jason
> > Can you check what the difference is between the two servers after you
> > kill the copy? Do they differ in the number of files?
> > Krishna
> >
>It's alright. I figured it out - there are major differences between
>Fedora9 and CentOS5 - which means du and find return different values!
>In the end I did a
>"rsync -avn /home/export/ remote:/home/export" and THAT reported that
>everything was in sync - so it is working. As to why du/etc returned
>different values? I dunno. The CentOS5 server is running sw-raid and the
>FC9 is an LVM encrypted volume - but I wouldn't have thought that would
>change blocksizes/whatever. Doesn't matter - glusterfs worked! :-)
>
>--
>Cheers
>
>Jason Haar
>Information Security Manager, Trimble Navigation Ltd.
>Phone: +64 3 9635 377 Fax: +64 3 9635 417
>PGP Fingerprint: 7A2E 0407 C9A6 CAF6 2B9F 8422 C063 5EBB FE1D 66D1
>
>
>_______________________________________________
>Gluster-users mailing list
>Gluster-users at gluster.org
>http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users





More information about the Gluster-users mailing list