Hi Patrick,
Thank you for your response. I read that the default was 1024K and that
there are a few approaches that can affect the performance. I read that
here http://nfs.sourceforge.net/nfs-howto/ar01s05.html
Furthermore I see the following when I use `dd` as suggested on the page
when writing a chunks of 16K. The result is:
$ time dd if=/dev/zero of=testfile bs=16k count=16384
16384+0 records in
16384+0 records out
268435456 bytes (268 MB) copied, 7.4459 s, 36.1 MB/s
real 0m7.447s
user 0m0.001s
sys 0m0.195s
The following secction on the page suggests to increase memory limits
5.7. Memory Limits on the Input Queue
....
....
# echo 262144 > /proc/sys/net/core/rmem_default
# echo 262144 > /proc/sys/net/core/rmem_max
....
....
My current settings are:
$ cat /proc/sys/net/core/rmem_default
229376
$ cat /proc/sys/net/core/rmem_max
229376
This line here:
rw,relatime,vers=3,rsize=8192,wsize=8192,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.6.9,mountvers=3,mountport=35079,mountproto=udp,fsc,local_lock=none,addr=10.0.6.9
is handled by autofs which is configured through (FreeIPA).
I could increase the wsize and rsize though but not sure what value to
define.
So for example if I do a `dd` using 8K chunks, the result is:
$ time dd if=/dev/zero of=testfile2 bs=8k count=16384
16384+0 records in
16384+0 records out
134217728 bytes (134 MB) copied, 3.0388 s, 44.2 MB/s
real 0m3.040s
user 0m0.001s
sys 0m0.098s
And here is the result using 64K chunks:
$ time dd if=/dev/zero of=testfile2 bs=64k count=16384
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 30.8299 s, 34.8 MB/s
real 0m30.933s
user 0m0.000s
sys 0m0.546s
On 06/03/2016 04:18 PM, Dave Howorth wrote:
> On 2016-06-03 14:49, Patrick J. LoPresti wrote:
>>> /Users/users from homeserver.internal.local:/users/users
>>> Flags:
>>>
>> rw,relatime,vers=3,rsize=8192,wsize=8192,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.6.9,mountvers=3,mountport=35079,mountproto=udp,fsc,local_lock=none,addr=10.0.6.9
>>
>>
>> How did you wind up with rsize/wsize of 8K? The default on Linux has
>> been 1
>> megabyte for a long time.
>
> Do you have a link for that, please? What is 'a long time'? Does it
> apply to all NFS versions and v3 in particular as the OP asked? Is there
> a list somewhere of NFS and kernel versions and defaults?
>
> Cheers, Dave
>
>> Fix this, and I suspect you will find performance increases dramatically.
>>
>> - Pat
>>
--
Valentin Bajrami
Kapteyn Astronomical Institute
University of Groningen
Postbus 800
NL-9700 AV Groningen
The Netherlands
Phone: +31-(0)50-3634068
|
