Louwrentius

TL;DR Yes it works with some modifications of the driver source.

A USB Wi-Fi adapter I used with a Raspberry Pi broke as I dropped it on the floor, so I had to replace it. I just went to a local shop and bought the Sitecom AC600 adapter as that's what they had available (with support for 5Ghz networking).

I had some hope that I would just plug it in and it would 'just work™'. But no. Linux. In the end, the device cost me 30 euro's including taxes, but the time spend to get it to work may have made this a very expensive USB Wi-Fi dongle. And it's funny to think about the fact that the Wi-Fi dongle is almost the same price as the Raspberry Pi board itself.

But I did get it working and I'd like to show you how.

It started with a google for 'sitecom ac600 linux' which landed me on this page. This page told me the device uses a MediaTek chipset (MT7610U).

So you need to download the driver from MediaTek. Here is a direct link

So you may do something like this:

cd /usr/src
wget http://s3.amazonaws.com/mtk.cfs/Downloads/linux/mt7610u_wifi_sta_v3002_dpo_20130916.tar.bz2
tar xjf mt7610u_wifi_sta_v3002_dpo_20130916.tar.bz2
cd mt7610u_wifi_sta_v3002_dpo_20130916

Now you would hope that it's just like this:

make
make install

And we're happy right? Linux FTW! Well, NO! We're using Linux so we have to work for stuff that works right out of the box on Windows and Mac OS.

So we first start with editing "include/os/rt_linux.h" and go to line ~279. There we make sure that we edit the struct like this:

    typedef struct _OS_FS_INFO_
 {
    kuid_t              fsuid;
    kgid_t              fsgid;
    mm_segment_t    fs;
 } OS_FS_INFO;

Basically, the words int are replaced by kuid_t and kgid_t, or else, compilation will abort with an error.

Ofcourse, the Sitecom AC600 has an USB identifier that is unknown to the driver, so after compilation, it still doesn't work.

lsusb output:

Bus 001 Device 004: ID 0df6:0075 Sitecom Europe B.V.

So google landed me on this nice thread by 'praseodym' that explained the remaining steps. I stole the info below from this thread.

So while we are in the source directory of the module, we are going to edit "common/rtusb_dev_id.c" and add

{USB_DEVICE(0x0DF6,0x0075)}, /* MT7610U */

So this will make the AC600 gets recognised by the driver. Now we also need to edit "os/linux/confik.mk" and change these lines like this:

HAS_WPA_SUPPLICANT=y
HAS_NATIVE_WPA_SUPPLICANT_SUPPORT=y

So no, we are still not ready yet. I'm not 100 percent sure that this is required anymore, but I found this nice thread in Italian and a very small comment by 'shoe rat' tucked away at the end that may make the difference between a working device or not.

We need to edit the file "os/linux/config.mk" and go to line ~663. Then, around that line, change

CHIPSET_DAT = 2860

to:

CHIPSET_DAT = 2870

Yes. Finally! Now you can do:

make
make install

Imagine that such a 'make' takes about 20 minutes on a Raspbery Pi. No joke.

Now you can either do this:

modprobe mt7650u_sta

You should see something like this:

root@raspberrypi:/usr/src# lsmod
Module                  Size  Used by
snd_bcm2835            16181  0 
snd_pcm                63684  1 snd_bcm2835
snd_page_alloc          3604  1 snd_pcm
snd_seq                43926  0 
snd_seq_device          4981  1 snd_seq
snd_timer              15936  2 snd_pcm,snd_seq
snd                    44915  5 snd_bcm2835,snd_timer,snd_pcm,snd_seq,snd_seq_device
soundcore               4827  1 snd
mt7650u_sta           895786  1 
pl2303                  7951  0 
usbserial              19536  1 pl2303

You should be able to see a 'ra0' device when entering ifconfig -a or iwconfig and just configure it like any wireless device (out-of-scope).

So once up-and-running, the Sitecom AC600 works fine under Linux and even sees and connects to 5 GHz networks. But not without a caveat of-course. I needed to configure a 5 GHz channel below 100 (I chose 48) on my Apple Airport Extreme, or the Wi-Fi dongle would not see the 5GHz network and would not be able to connect to it.

So I hope somebody else is helped by this information.

At work, access to the internet is provided by a 10 Mbit down / 1 Mbit up ADSL-connection. As we are a mid-size company, bandwidth is clearly a severe constraint. But it was not our biggest problem. Even simple web-browsing was very slow.

As I was setting up a monitoring environment based on Nagios and pnp4nagios, I started to graph the latency of our internet connection, just to prove that we have a problem.

Boy did we get that proof:

Just look at the y-axis, which scale is in milliseconds. For most of the day, the average latency is 175 ms, with some high spikes. Just browsing the web was a pain during times of high-latency, which was clearly almost all of the time.

I became so fed up with our slow internet access that I decided to take matters in my own hands and resolve the low latency issue. The solution? Traffic shaping.

I learned that as ADSL-connections are saturated, especially their upload capacity, you will experience high latency and packet loss. So the trick is to never saturate the connection.

I grabbed a Linux box with two network interfaces and placed it between our internet router and our firewall in bridge mode.

For actual traffic shaping I used wondershaper which is part of Debian or Ubuntu.

apt-get install wondershaper

The wondershaper script is extremely simple, it's specifically build to resolve the problem we face with our ADSL connection. It not only prioritises traffic, it allows you to limit bandwidth usage and thus prevent you from saturating the connection.

This simple example limits bandwidth a bit below full capacity, which dramatically improved latency.

Syntax:

wondershaper <interface> <rx> <tx>

Example:

wondershaper eth1 9500 700

As you can see, latency improved dramatically:

Again, look at the y-axis. We went from an average latency of 175 ms to an average of 35 ms. That's quite an improvement.

Can you spot on which day I implemented traffic shaping?

At the time of writing this blog post, the company is working on fiber internet access, resolving our internet woes, but it will take quite some time before that will be installed, so this is a nice intermediate solution.

Linux Bonding

In this article I'd like to show the results of using regular 1 Gigabit network connections to achieve 450 MB/s file transfers over NFS.

I'm again using Linux interface bonding for this purpose.

Linux interface bonding can be used to create a virtual network interface with the aggregate bandwidth of all the interfaces added to the bond. Two gigabit network interfaces will give you - guess what - two gigabit or ~220 MB/s of bandwidth.

This bandwidth can be used by a single TCP-connection.

So how is this achieved? The Linux bonding kernel module has special bonding mode: mode 0 or round-robin bonding. In this mode, the kernel will stripe packets across the interfaces in the 'bond' like RAID 0 with hard drives. As with RAID 0, you get additional performance with each device you add.

So I've added HP NC364T quad-port network cards to my servers. Thus each server has a theoretical bandwidth of 4 Gigabit. These HP network cards cost just 145 Euro and I even found the card for 105 Dollar on Amazon.

With two servers, you just need four UTP cable's to connect the two network interfaces and you're done. This would cost you ~300 Euro or ~200 dollar in total.

If you want to connect additional servers, you need a managed gigabit switch with VLAN-support and sufficient ports. Each additional server will use 4 ports on the switch, excluding interfaces for remote access and other purposes.

Managed gigabit switches are quite inexpensive these days. I bought a 24 port switch: HP 1810-24G v2 (J9803A) for about 180 euros (209 Dollars on Newegg) and it can even be rack-mounted.

Using VLANS

So you can't just use a gigabit switch and connect all these quad-port network cards to a single VLAN. I tested this scenario first and only got a maximum transfer speed of 270 MB/s while copying a file between servers over NFS.

The trick is to create a separate VLAN for every network port. So if you use a quad-port network card, you need four VLANs. Also, you must make sure that every port on every network card is in the same VLAN. For example, port 1 on every card needs to be in VLAN 21, port 2 in VLAN 22, and so on. You also must add the appropriate switch port to the correct VLAN. Last, you must add the network interfaces to the bond in the right order.

So why do you need to use VLANs? The reason is quite simple. Bonding works by spoofing the same hardware or MAC-address on all interfaces. So the switch sees the same hardware address on four ports, and thus gets confused. To which port should the packet be sent?

If you put each port in it's own VLAN, the 'spoofed' MAC-address is seen only once in each VLAN. So the switch won't be confused. What you are in fact doing by creating VLANs is creating four separate switches. So if you have - for example - four cheap 8-port unmanaged gigabit switches, this would work too.

So assuming that you have four ethernet interfaces, this is an example of how you can create the bond:

ifenslave bond0 eth1 eth2 eth3 eth4

Next, you just assign an IP-address to the bond0 interface, like you would with a regular eth(x) interface.

ifconfig bond0 192.168.2.10 netmask 255.255.255.0

Up to this point, I only achieved about 350 MB/s. I needed to enable jumbo frames on all interfaces and the switch to achieve 450 MB/s.

ifconfig bond0 mtu 9000 up

Next, you can just mount any NFS share over the interface and start copying files. That's all.

Once I added each interface to the appropriate VLAN, I got about 450 MB/s for a single file copy with 'cp' over NFS.

I did not perform a 'cp' but a 'dd' because I don't have a disk array fast enough (yet) that can write at 450 MB/s.

So for three servers, this solution will cost me 600 Euro or 520 Dollar.

What about LAGs and LACP?

Don't configure your clients or switches for LACP, it doesn't give you the speed benefit and it's not required.

10 Gigabit ethernet?

Frankly, it's quite expensive if you need to connect more than two servers. An entry-level 10Gbe NIC like the Intel X540-T1 does about 300 Euros or 450 Dollar. This card allows you to use Cat 6e UTP Cabling. (Pricing from Dutch Webshops in Euros and Newegg in Dollars).

With two of those, you can have 10Gbit ethernet between two servers for 600 Euro or 700 Dollar. If you need to connect more servers, you would need a switch. The problem is that 10Gbit switches are not cheap. An 8-port unmanaged switch from Netgear (ProSAFE Plus XS708E) does about 720 Euro's or 900 Dollar.

If you want to connect three servers, you need three network cards and a switch. So three network cards and a switch will cost you 900 Euro (1050 Dollar) for the network cards and 720 Euro (900 Dollar) for the switch, totalling 1800 euro or 1950 Dollar.

You will get higher transfer speeds, but at a significantly higher price.

For many business purposes this higher price can be easily justified and I would select 10 Gb over 1 Gb bonding in a heart-beat. Less cables, higher performance, lower latency.

However, bonding gigabit interfaces allows you to use off-the-shelf equipment and maybe a nice compromis between cost, usability and performance.

Operating System Support

As far as I'm aware, round-robin bonding is only supported on Linux. Other operating systems do not support it.