3 replies to this topic

[bleskone]

hello,

Is this feature enabled in new kernel-image 4.10 ?
It was allowed in previous kernel 4.7


root@hd51:~# ifconfig eth0 mtu 9000
SIOCSIFMTU: Invalid argument

used in multiple recording to nfs share through udp protocol

thanks for support

[Erik Slagter]

It has nothing to do with kernel support, the STB ethernet interfaces are the problem (hardware). They're so simple you wouldn't get an improvement from increasing the MTU anyway.

[bleskone]

thanks for reply,

yes, it looks like hardware limitation, but in previous kernel 4.7, I could setup bigger mtu, not mentioning that it was not very stable and box freezes after some time.

when using nfs over TCP, overhead and iowait of cpu of nfs server(cca 10%) is quite high, when using nfs over udp with mtu 1500, fragmentation of packets is encoutered, and also iowait is high, running 4 concurrent recordings to nfs share

from man nfs:
Using NFS over UDP on high-speed links
Using NFS over UDP on high-speed links such as Gigabit can cause silent
data corruption.

The problem can be triggered at high loads, and is caused by problems
in IP fragment reassembly. NFS read and writes typically transmit UDP
packets of 4 Kilobytes or more, which have to be broken up into several
fragments in order to be sent over the Ethernet link, which limits
packets to 1500 bytes by default. This process happens at the IP net‐
work layer and is called fragmentation.
...

If you absolutely have to use NFS over UDP over Gigabit Ethernet, some
steps can be taken to mitigate the problem and reduce the probability
of corruption:

Jumbo frames: Many Gigabit network cards are capable of transmitting
frames bigger than the 1500 byte limit of traditional
Ethernet, typically 9000 bytes. Using jumbo frames of
9000 bytes will allow you to run NFS over UDP at a page
size of 8K without fragmentation. Of course, this is
only feasible if all involved stations support jumbo
frames.

[Erik Slagter]

These hints are only applicable on "normal" ethernet interfaces.