Pedro Martelletto wrote:
On Sun, Nov 12, 2006 at 01:59:47PM +0100, Federico Giannici wrote:
Is there any case that makes NFS in 4.0 read the listing of a directory?
Yes, the getcwd() change. I wonder if it exposed any other bug in our
NFS code (as it did in the past, but those got fixed, since they were
reported).
Anyway, I'm working on trying to find a reason for what you're seeing,
along with a fix.
I have just noticed that at the same time of these NFS problems there
has been a big decrease of traffic in our DNS server...
After some investigation I found that the DNS server (recently upgraded
to 4.0) had reached the maximum of memory (there was a "datasize 200M"
option).
Now that the DNS problem is solved, it SEEMS that the problem with NFS
is reduced. So I have two questions:
1) Is NFS activity in some way related to DNS?
Anyway it could be that my "mail delivery" program depends on DNS, and
so it delayed the delivery...
2) Is it possible that, if a certain number of processes are already
using NFS, subsequent attempts by other processes is STOPPED, until some
other process RELEASE the use of NFS???
If this is true, what I can do to increase the number of concurrent
processes writing via NFS?
I increased "vfs.nfs.iothreads" to 20, but it seems not be enough...
Thanks.
--
___________________________________________________
__
|- [EMAIL PROTECTED]
|ederico Giannici http://www.neomedia.it
___________________________________________________
