Re: Performance of Wireguard on Infiniband 40G

[Jason A. Donenfeld]

Hey again,

On Sat, May 13, 2017 at 9:37 AM, Baptiste Jonglez
 wrote:
> - Debian jessie
> - Linux 3.16.43-2

One small and unfortunate thought just occurred to me: the backporting
to really old kernels I'm pretty sure is way less efficient than newer
kernels on the RX, due to some missing core fast-path APIs in the old
kernels. In particular, I had to wrap the UDP layer with some nasty
hacks to get packets out, whereas newer kernels have an elegant API
for that which integrates in the right place. Just a thought... I
haven't actually done concrete measurements though.

Jason
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

Re: TX counters not updated on wireguard interface

[Jason A. Donenfeld]

Okay I figured it out. Upstream
039f50629b7f860f36644ed1f34b27da9aa62f43 only came in 4.5, so explicit
calls to iptunnel_xmit_stats is required. I'll play with this and see
if I can fix it.

Jason
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

Re: TX counters not updated on wireguard interface

[Jason A. Donenfeld]

Hey Baptiste,

I'll try to reproduce this shortly. I did make a change to some compat
code [1] for kernels older than 3.12 to fix a similar bug. But Jesse
has a newer kernel than that, I think? I'll check into it and try to
fix. Thanks for reporting.

Jason

[1] 
https://git.zx2c4.com/WireGuard/patch/?id=1370d96432c2c7e86705ce99b1682a3b7c2ce1cd
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

Re: Performance of Wireguard on Infiniband 40G

[Jason A. Donenfeld]

Hey Baptiste,

Awesome test! Thanks for reporting the results.

On Sat, May 13, 2017 at 9:37 AM, Baptiste Jonglez
 wrote:
> Using iperf (TCP mode) over the wireguard interface, performance was
> around 1.6 Gbit/s.  In bidirectional mode (iperf -d), performance was
> 700 Mbit/s + 800 Mbit/s.

Indeed the current multicore algorithm has a lot of issues. Samuel,
CCd, is going to be doing some work on optimizing this algorithm this
summer.

> After raising the MTU of the wireguard interface to 65450, performance
> went up to 7.6 Gbit/s (unidirectional iperf).

It makes sense that it'd be higher, since CPUs work best when running
uninterrupted, but still this indicates that padata is a very
suboptimal algorithm. Expect some improvements on this in the coming
months. Hopefully you'll be able to test on similar hardware at some
point when things are finished.

> Note that infiniband has a MTU of 65520 bytes, but Wireguard still selects
> a MTU of 1420 bytes for its interface.

Yea the 1420 is just a hard coded "default". I probably add something
clever to autoselect an MTU when configuring the first peer's first
endpoint (by computing the route and taking its interface's mtu and
doing subtraction, etc), but the long term solution, I think, will be
to do some more clever PMTU situation from within WireGuard. I'm still
working out exactly how to do this, but it should be possible.


> - Xeon E5520 @2.27GHz (2 CPUs, 4 cores each)
> - Mellanox ConnectX IB 4X QDR MT26428

*drools* That's some awesome hardware!

> - iperf 2.0.5

iperf2 has the -b bidirectional mode which is nice, but it seems like
most people are using iperf3 now. Out of curiosity, is there a reason
for preferring iperf2, beyond the -b switch?

Jason
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

Performance of Wireguard on Infiniband 40G

[Baptiste Jonglez]

Hi,

Just for information, I did a quick test of Wireguard over a 40G
Infiniband network, between two machines with a Xeon E5520.

Using iperf (TCP mode) over the wireguard interface, performance was
around 1.6 Gbit/s.  In bidirectional mode (iperf -d), performance was
700 Mbit/s + 800 Mbit/s.

Note that infiniband has a MTU of 65520 bytes, but Wireguard still selects
a MTU of 1420 bytes for its interface.
After raising the MTU of the wireguard interface to 65450, performance
went up to 7.6 Gbit/s (unidirectional iperf).

Using the Infiniband network directly, iperf's performance is 21.7 Gbit/s
(iperf maxes out the CPU at the receiver, even when using 8 threads).

Hardware used:

- Xeon E5520 @2.27GHz (2 CPUs, 4 cores each)
- Mellanox ConnectX IB 4X QDR MT26428

Versions used:

- Debian jessie
- Linux 3.16.43-2
- Wireguard 0.0.20170421-2
- iperf 2.0.5
- Mellanox ConnectX InfiniBand driver v2.2-1

Baptiste


signature.asc
Description: PGP signature
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

Rebasing Wireguard's master branch

[Baptiste Jonglez]

Hi Jason,

Could you please stop rebasing the master branch of the git repository?
It's really annoying (and possibly confusing) to obtain merge conflicts
while simply pulling the latest changes.

If you want to rebase, please use feature branches or something similar.

Thanks,
Baptiste


signature.asc
Description: PGP signature
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

TX counters not updated on wireguard interface

[Baptiste Jonglez]

Hi,

While trying 0.0.20170421-2 on Debian jessie, I noticed that the TX
counters are not updated:

$ ip -s link show
7: wg:  mtu 1420 qdisc noqueue state UNKNOWN 
mode DEFAULT group default
link/none
RX: bytes  packets  errors  dropped overrun mcast
4384059324 4073303  0   0   0   0
TX: bytes  packets  errors  dropped carrier collsns
0  00   0   0   0

This is after a bidirectional TCP iperf, so they has been some trafic in
both directions.  Bandwidth monitors like bmon also don't show any
outgoing trafic.

Baptiste


signature.asc
Description: PGP signature
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard

Installation instructions for Debian

[Baptiste Jonglez]

Hi Jason,

Could you update the installation instructions for Debian [1]?

Based on the discussion from a few months ago, and given that Wireguard
now supports Jessie's 3.16 kernel, these instructions should be added:

For Debian jessie or stretch:
# echo "deb http://deb.debian.org/debian/ unstable main" > 
/etc/apt/sources.list.d/unstable-wireguard.list
# cat < /etc/apt/preferences.d/limit-unstable
Package: *
Pin: release a=unstable
Pin-Priority: 200
EOF
# apt update
# apt install wireguard-dkms wireguard-tools

I have verified that the resulting setup is functional (with 0.0.20170421-2 on 
jessie).

Thanks,
Baptiste

[1] https://www.wireguard.io/install/


signature.asc
Description: PGP signature
___
WireGuard mailing list
WireGuard@lists.zx2c4.com
https://lists.zx2c4.com/mailman/listinfo/wireguard