Hello,
It would be a good idea to update the Documentation about CIFS:
http://gluster.com/community/documentation/index.php/Gluster_3.2:_Exporting_Gluster_Volumes_Through_Samba
When the simple truth is, that there is no CIFS Support in gluster itself - it should not be in the
docs.
As I
On OSS server i using ext4.
And I presume you are using Linux on the OSS servers ?
Yes. I'm using CentOS 5.5 x86_64.
С уважением,
Алексанян Александр
ОАО Т-Платформы
Тел: +7(495)744-0980 (1434)
От: Pavan [t...@gluster.com]
Отправлено: 4 мая 2011
How can I prevent client to querying all glusterfs servers for updated
version of file and use version in a cache instead? I am using
glusterfs for webserving and even with opcode caching there are lots
of stat calls (it is nature of webapp I am running). I now myself when
I am updating code files
I had very same problems as you do, disabling stat-prefetch helped a
lot, but sometimes problem with constant self healing occur and in
that case if you run md5sum on files you can find that they are differ
on bricks. Disable stat-prefetch and you'll make your life easier,
but not hasslefree (
What exactly is stat-prefetch and how does it help?
On Thu, May 5, 2011 at 4:27 PM, Max Ivanov ivanov.ma...@gmail.com wrote:
I had very same problems as you do, disabling stat-prefetch helped a
lot, but sometimes problem with constant self healing occur and in
that case if you run md5sum on
On Thu, May 05, 2011 at 05:26:21PM -0700, Mohit Anchlia wrote:
What exactly is stat-prefetch and how does it help?
It's directory caching, so it can make ls -l results appear faster for
instance. There is a comparative lag with that the first time into a
directory if it's turned off.
It's