You can easily configure your own cloud-init configuration into your LXD instance profile.
In my case, I want cloud-init to automatically ssh-import-id kirkland, to fetch my keys from Launchpad. Alternatively, I could use gh:dustinkirkland to fetch my keys from Github.
Here's how!
First, edit your default LXD profile (or any other, for that matter):
I've been one of DD-WRT's biggest fans, for more than 10 years. I've always flashed my router with custom firmware, fine-tuned my wired and wireless networks, and locked down a VPN back home. I've genuinely always loved tinkering with network gear.
A couple of weeks ago, I decided to re-deploy my home network. I've been hearing about Ubiquiti Networks from my colleagues at Canonical, where we use Ubiquiti gear for our many and varied company events. Moreover, it seems a number of us have taken to running the same kits in our home offices.
There's something quite unique about the UniFi Controller -- the server that "controls" your router, gateway, and access points. Rather than being built into the USG itself, you run the server somewhere else.
Sure you can buy their hardware appliance (which I'm sure is nice). But you can just as easily run it on an Ubuntu machine yourself. That machine could be a physical machine on your network, a virtual machine locally or in the cloud, or it could be an LXD machine container.
I opted for the latter. I'm happily running the UniFi Controller in a LXD machine container, and it's easy for you to setup, too.
I'm running Ubuntu 16.04 LTS 64-bit on an Intel NUC somewhere in my house. It happens to be running Ubuntu Desktop, as it's attached to one of the TVs in my house, as a media playing device. In it's spare time, it's a server I use for LXD, Docker, and other development purposes.
I've configured the network on the machine to "bridge" LXD to my USG router, which happens to be running DHCP and DNS. I'm going to move that to a MAAS server, but that's a post for another day.
Here's /etc/network/interfaces on that machine:
kirkland@masterbr:~⟫ cat /etc/network/interfaces
# interfaces(5) file used by ifup(8) and ifdown(8)
auto lo
iface lo inet loopback
auto eth0
iface eth0 inet manual
auto br0
iface br0 inet dhcp
bridge_ports eth0
bridge_stp off
bridge_fd 0
bridge_maxwait 0
So eth0 is bridged, to br0. ifconfig looks like this:
It's important to notice that this container drew an IP address on my 10.0.0.0/24 LAN. It will need this, to detect, federate, and manage the Ubiquiti hardware.
Now, let's exec into it, and import our SSH keys, so that we can SSH into it later.
kirkland@masterbr:~⟫ lxc exec unifi-controller bash
root@unifi-controller:~# ssh-import-id kirkland
2016-12-09 21:56:36,558 INFO Authorized key ['4096', 'd3:dd:e4:72:25:18:f3:ea:93:10:1a:5b:9f:bc:ef:5e', 'kirkland@x220', '(RSA)']
2016-12-09 21:56:36,568 INFO Authorized key ['2048', '69:57:f9:b6:11:73:48:ae:11:10:b5:18:26:7c:15:9d', 'kirkland@mac', '(RSA)']
2016-12-09 21:56:36,569 INFO [2] SSH keys [Authorized]
root@unifi-controller:~# exit
exit
kirkland@masterbr:~⟫ ssh root@10.0.0.183
The authenticity of host '10.0.0.183 (10.0.0.183)' can't be established.
ECDSA key fingerprint is SHA256:we0zAxifd0dcnAE2tVE53NFbQCop61f+MmHGsyGj0Xg.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '10.0.0.183' (ECDSA) to the list of known hosts.
root@unifi-controller:~#
Now, let's add the Unifi repository and install the deb and all its dependencies. It's a big pile of Java and MongoDB, which I'm happy to keep nicely "contained" in this LXD instance!
root@unifi-controller:~# echo deb http://www.ubnt.com/downloads/unifi/debian stable ubiquiti
deb http://www.ubnt.com/downloads/unifi/debian stable ubiquiti
root@unifi-controller:~# echo "deb http://www.ubnt.com/downloads/unifi/debian stable ubiquiti" | sudo tee -a /etc/apt/sources.list
deb http://www.ubnt.com/downloads/unifi/debian stable ubiquiti
root@unifi-controller:~# apt-key adv --keyserver keyserver.ubuntu.com --recv C0A52C50
Executing: gpg --ignore-time-conflict --no-options --no-default-keyring --homedir /tmp/tmp.hhgdd0ssJQ --no-auto-check-trustdb --trust-model always --keyring /etc/apt/trusted.gpg --primary-keyring /etc/apt/trusted.gpg --keyserver keyserver.ubuntu.com --recv C0A52C50
gpg: requesting key C0A52C50 from hkp server keyserver.ubuntu.com
gpg: key C0A52C50: public key "UniFi Developers " imported
gpg: Total number processed: 1
gpg: imported: 1 (RSA: 1)
root@unifi-controller:~# apt update >/dev/null 2>&1
root@unifi-controller:~# apt install unifi
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following package was automatically installed and is no longer required:
os-prober
Use 'apt-get autoremove' to remove it.
The following extra packages will be installed:
binutils ca-certificates-java default-jre-headless fontconfig-config
fonts-dejavu-core java-common jsvc libasyncns0 libavahi-client3
libavahi-common-data libavahi-common3 libboost-filesystem1.54.0
libboost-program-options1.54.0 libboost-system1.54.0 libboost-thread1.54.0
libcommons-daemon-java libcups2 libflac8 libfontconfig1 libgoogle-perftools4
libjpeg-turbo8 libjpeg8 liblcms2-2 libnspr4 libnss3 libnss3-nssdb libogg0
libpcrecpp0 libpcsclite1 libpulse0 libsctp1 libsnappy1 libsndfile1
libtcmalloc-minimal4 libunwind8 libv8-3.14.5 libvorbis0a libvorbisenc2
lksctp-tools mongodb-clients mongodb-server openjdk-7-jre-headless tzdata
tzdata-java
Suggested packages:
binutils-doc default-jre equivs java-virtual-machine cups-common
liblcms2-utils pcscd pulseaudio icedtea-7-jre-jamvm libnss-mdns
sun-java6-fonts fonts-dejavu-extra fonts-ipafont-gothic fonts-ipafont-mincho
ttf-wqy-microhei ttf-wqy-zenhei ttf-indic-fonts-core ttf-telugu-fonts
ttf-oriya-fonts ttf-kannada-fonts ttf-bengali-fonts
The following NEW packages will be installed:
binutils ca-certificates-java default-jre-headless fontconfig-config
fonts-dejavu-core java-common jsvc libasyncns0 libavahi-client3
libavahi-common-data libavahi-common3 libboost-filesystem1.54.0
libboost-program-options1.54.0 libboost-system1.54.0 libboost-thread1.54.0
libcommons-daemon-java libcups2 libflac8 libfontconfig1 libgoogle-perftools4
libjpeg-turbo8 libjpeg8 liblcms2-2 libnspr4 libnss3 libnss3-nssdb libogg0
libpcrecpp0 libpcsclite1 libpulse0 libsctp1 libsnappy1 libsndfile1
libtcmalloc-minimal4 libunwind8 libv8-3.14.5 libvorbis0a libvorbisenc2
lksctp-tools mongodb-clients mongodb-server openjdk-7-jre-headless
tzdata-java unifi
The following packages will be upgraded:
tzdata
1 upgraded, 44 newly installed, 0 to remove and 10 not upgraded.
Need to get 133 MB of archives.
After this operation, 287 MB of additional disk space will be used.
Do you want to continue? [Y/n] y
...
done.
Finally, we point a web browser at this server, http://10.0.0.183:8443/ in my case, and run through the UniFi setup there.
A couple of weeks ago, I delivered a talk at the Container Camp UK 2016. It was an brilliant event, on a beautiful stage at Picturehouse Central in Picadilly Circus in London.
I had the opportunity to speak at Container World 2016 in Santa Clara yesterday. Thanks in part to the Netflix guys who preceded me, the room was absolutely packed!
Ubuntu 16.04 LTS (Xenial) is only a few short weeks away, and with it comes one of the most exciting new features Linux has seen in a very long time...
ZFS -- baked directly into Ubuntu -- supported by Canonical.
What is ZFS?
ZFS is a combination of a volume manager (like LVM) and a filesystem (like ext4, xfs, or btrfs).
ZFS one of the most beloved features of Solaris, universally coveted by every Linux sysadmin with a Solaris background. To our delight, we're happy to make to OpenZFS available on every Ubuntu system. Ubuntu's reference guide for ZFS can be found here, and these are a few of the killer features:
snapshots
copy-on-write cloning
continuous integrity checking against data corruption
automatic repair
efficient data compression.
These features truly make ZFS the perfect filesystem for containers.
What does "support" mean?
You'll find zfs.ko automatically built and installed on your Ubuntu systems. No more DKMS-built modules!
Now, let's install lxd and zfsutils-linux, if you haven't already:
$ sudo apt install lxd zfsutils-linux
Next, let's use the interactive lxd init command to setup LXD and ZFS. In the example below, I'm simply using a sparse, loopback file for the ZFS pool. For best results (and what I use on my laptop and production servers), it's best to use a raw SSD partition or device.
$ sudo lxd init
Name of the storage backend to use (dir or zfs): zfs
Create a new ZFS pool (yes/no)? yes
Name of the new ZFS pool: lxd
Would you like to use an existing block device (yes/no)? no
Size in GB of the new loop device (1GB minimum): 2
Would you like LXD to be available over the network (yes/no)? no
LXD has been successfully configured.
We can check our ZFS pool now:
$ sudo zpool list
NAME SIZE ALLOC FREE EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
lxd 1.98G 450K 1.98G - 0% 0% 1.00x ONLINE -
$ sudo zpool status
pool: lxd
state: ONLINE
scan: none requested
config:
NAME STATE READ WRITE CKSUM
lxd ONLINE 0 0 0
/var/lib/lxd/zfs.img ONLINE 0 0 0
errors: No known data errors
$ lxc config get storage.zfs_pool_name
storage.zfs_pool_name: lxd
Finally, let's import the Ubuntu LXD image, and launch a few containers. Note how fast containers launch, which is enabled by the ZFS cloning and copy-on-write features:
$ newgrp lxd
$ lxd-images import ubuntu --alias ubuntu
Downloading the GPG key for http://cloud-images.ubuntu.com
Progress: 48 %
Validating the GPG signature of /tmp/tmpa71cw5wl/download.json.asc
Downloading the image.
Image manifest: http://cloud-images.ubuntu.com/server/releases/trusty/release-20160201/ubuntu-14.04-server-cloudimg-amd64.manifest
Image imported as: 54c8caac1f61901ed86c68f24af5f5d3672bdc62c71d04f06df3a59e95684473
Setup alias: ubuntu
$ for i in $(seq 1 5); do lxc launch ubuntu; done
...
$ lxc list
+-------------------------+---------+-------------------+------+-----------+-----------+
| NAME | STATE | IPV4 | IPV6 | EPHEMERAL | SNAPSHOTS |
+-------------------------+---------+-------------------+------+-----------+-----------+
| discordant-loria | RUNNING | 10.0.3.130 (eth0) | | NO | 0 |
+-------------------------+---------+-------------------+------+-----------+-----------+
| fictive-noble | RUNNING | 10.0.3.91 (eth0) | | NO | 0 |
+-------------------------+---------+-------------------+------+-----------+-----------+
| interprotoplasmic-essie | RUNNING | 10.0.3.242 (eth0) | | NO | 0 |
+-------------------------+---------+-------------------+------+-----------+-----------+
| nondamaging-cain | RUNNING | 10.0.3.9 (eth0) | | NO | 0 |
+-------------------------+---------+-------------------+------+-----------+-----------+
| untreasurable-efrain | RUNNING | 10.0.3.89 (eth0) | | NO | 0 |
+-------------------------+---------+-------------------+------+-----------+-----------+
There's no shortage of excitement, controversy, and readership, any time you can work "Docker" into a headline these days. Perhaps a bit like "Donald Trump", but for CIO tech blogs and IT news -- a real hot button. Hey, look, I even did it myself in the title of this post!
Docker's default container image is certainly Docker's decision to make. But it would be prudent to examine at a few facts:
(1) Check DockerHub and you may notice that while Busybox (Alpine Linux) has surpassed Ubuntu in the number downloads (66M to 40M), Ubuntu is still by far the most "popular" by number of "stars" -- likes, favorites, +1's, whatever, (3.2K to 499).
(2) Ubuntu's compressed, minimal root tarball is 59 MB, which is what is downloaded over the Internet. That's different from the 188 MB uncompressed root filesystem, which has been quoted a number of times in the press.
(3) The real magic of Docker is such that you only ever download that base image, one time! And you only store one copy of the uncompressed root filesystem on your disk! Just once, sudo docker pull ubuntu, on your laptop at home or work, and then launch thousands of images at a coffee shop or airport lounge with its spotty wifi. Build derivative images, FROM ubuntu, etc. and you only ever store the incremental differences.
Actually, I encourage you to test that out yourself... I just launched a t2.micro -- Amazon's cheapest instance type with the lowest networking bandwidth. It took 15.938s to sudo apt install docker.io. And it took 9.230s to sudo docker pull ubuntu. It takes less time to download Ubuntu than to install Docker!
ubuntu@ip-172-30-0-129:~⟫ time sudo apt install docker.io -y
...
real 0m15.938s
user 0m2.146s
sys 0m0.913s
As compared to...
ubuntu@ip-172-30-0-129:~⟫ time sudo docker pull ubuntu
latest: Pulling from ubuntu
f15ce52fc004: Pull complete
c4fae638e7ce: Pull complete
a4c5be5b6e59: Pull complete
8693db7e8a00: Pull complete
ubuntu:latest: The image you are pulling has been verified. Important: image verification is a tech preview feature and should not be relied on to provide security.
Digest: sha256:457b05828bdb5dcc044d93d042863fba3f2158ae249a6db5ae3934307c757c54
Status: Downloaded newer image for ubuntu:latest
real 0m9.230s
user 0m0.021s
sys 0m0.016s
Now, sure, it takes even less than that to download Alpine Linux (0.747s by my test), but again you only ever do that once! After you have your initial image, launching Docker containers take the exact same amount of time (0.233s) and identical storage differences. See:
ubuntu@ip-172-30-0-129:/tmp/docker⟫ time sudo docker run alpine /bin/true
real 0m0.233s
user 0m0.014s
sys 0m0.001s
ubuntu@ip-172-30-0-129:/tmp/docker⟫ time sudo docker run ubuntu /bin/true
real 0m0.234s
user 0m0.012s
sys 0m0.002s
(4) I regularly communicate sincere, warm congratulations to our friends at Docker Inc, on its continued growth. shykes publicly mentioned the hiring of the maintainer of Alpine Linux in that Hacker News post. As a long time Linux distro developer myself, I have tons of respect for everyone involved in building a high quality Linux distribution. In fact, Canonical employs over 700 people, in 44 countries, working around the clock, all calendar year, to make Ubuntu the world's most popular Linux OS. Importantly, that includes a dedicated security team that has an outstanding track record over the last 12 years, keeping Ubuntu servers, clouds, desktops, laptops, tablets, and phones up-to-date and protected against the latest security vulnerabilities. I don't know personally Natanael, but I'm intimately aware of what a spectacular amount of work it is to maintain and secure an OS distribution, as it makes its way into enterprise and production deployments. Good luck!
(5) There are currently 5,854 packages available via apk in Alpine Linux (sudo docker run alpine apk search -v). There are 8,862 packages in Ubuntu Main (officially supported by Canonical), and 53,150 binary packages across all of Ubuntu Main, Universe, Restricted, and Multiverse, supported by the greater Ubuntu community. Nearly all 50,000+ packages are updated every 6 months, on time, every time, and we release an LTS version of Ubuntu and the best of open source software in the world every 2 years. Like clockwork. Choice. Velocity. Stability. That's what Ubuntu brings.
Docker holds a special place in the Ubuntu ecosystem, and Ubuntu has been instrumental in Docker's growth over the last 3 years. Where we go from here, is largely up to the cross-section of our two vibrant communities.
And so I ask you honestly...what do you want to see? How would you like to see Docker and Ubuntu operate together?
I'm Canonical's Product Manager for Ubuntu Server, I'm responsible for Canonical's relationship with Docker Inc, and I will read absolutely every comment posted below.
Cheers,
:-Dustin
p.s. I'm speaking at Container Summit in New York City today, and wrote this post from the top of the (inspiring!) One World Observatory at the World Trade Center this morning. Please come up and talk to me, if you want to share your thoughts (at Container Summit, not the One World Observatory)!
Picture yourself containers on a server With systemd trees and spawned tty's Somebody calls you, you answer quite quickly A world with the density so high
- Sgt. Graber's LXD Smarts Club Band
Last week, we proudly released Ubuntu 15.10 (Wily) -- the final developer snapshot of the Ubuntu Server before we focus the majority of our attention on quality, testing, performance, documentation, and stability for the Ubuntu 16.04 LTS cycle in the next 6 months.
Notably, LXD has been promoted to the Ubuntu Main archive, now commercially supported by Canonical. That has enabled us to install LXD by default on all Ubuntu Servers, from 15.10 forward.
That means that every Ubuntu server -- Intel, AMD, ARM, POWER, and even Virtual Machines in the cloud -- is now a full machine container hypervisor, capable of hosting hundreds of machine containers, right out of the box!
LXD in the Sky with Diamonds! Well, LXD is in the Cloud with Diamond level support from Canonical, anyway. You can even test it in your web browser here.
The development tree of Xenial (Ubuntu 16.04 LTS) has already inherited this behavior, and we will celebrate this feature broadly through our use of LXD containers in Juju, MAAS, and the reference platform of Ubuntu OpenStack, as well as the new nova-lxd hypervisor in the OpenStack Autopilot within Landscape.
While the young and the restless are already running Wily Ubuntu 15.10, the bold and the beautiful are still bound to their Trusty Ubuntu 14.04 LTS servers.
At Canonical, we understand both motivations, and this is why we have backported LXD to the Trusty archives, for safe, simple consumption and testing of this new generation of machine containers there, on your stable LTS.
Installing LXD on Trusty simply requires enabling the trusty-backports pocket, and installing the lxd package from there, with these 3 little commands:
In minutes, you can launch your first LXD containers. First, inherit your new group permissions, so you can execute the lxc command as your non-root user. Then, import some images, and launch a new container named lovely-rita. Shell into that container, and examine the process tree, install some packages, check the disk and memory and cpu available. Finally, exit when you're done, and optionally delete the container.
I was able to run over 600 containers simultaneously on my Thinkpad (x250, 16GB of RAM), and over 60 containers on an m1.small in Amazon (1.6GB of RAM).
Canonical is delighted to sponsor ContainerCon 2015, a Linux Foundation event in Seattle next week, August 17-19, 2015. It's quite exciting to see the A-list of sponsors, many of them newcomers to this particular technology, teaming with energy around containers.
From chroots to BSD Jails and Solaris Zones, the concepts behind containers were established decades ago, and in fact traverse the spectrum of server operating systems. At Canonical, we've been working on containers in Ubuntu for more than half a decade, providing a home and resources for stewardship and maintenance of the upstream Linux Containers (LXC) project since 2010.
Last year, we publicly shared our designs for LXD -- a new stratum on top of LXC that endows the advantages of a traditional hypervisor into the faster, more efficient world of containers.
LXD is a persistent daemon that provides a clean RESTful interface to manage (start, stop, clone, migrate, etc.) any of the containers on a given host.
Hosts running LXD are handily federated into clusters of container hypervisors, and can work as Nova Compute nodes in OpenStack, for example, delivering Infrastructure-as-a-Service cloud technology at lower costs and greater speeds.
Here, LXD and Docker are quite complementary technologies. LXD furnishes a dynamic platform for "system containers" -- containers that behave like physical or virtual machines, supplying all of the functionality of a full operating system (minus the kernel, which is shared with the host). Such "machine containers" are the core of IaaS clouds, where users focus on instances with compute, storage, and networking that behave like traditional datacenter hardware.
LXD runs perfectly well along with Docker, which supplies a framework for "application containers" -- containers that enclose individual processes that often relate to one another as pools of micro services and deliver complex web applications.
Moreover, the Zen of LXD is the fact that the underlying container implementation is actually decoupled from the RESTful API that drives LXD functionality. We are most excited to discuss next week at ContainerCon our work with Microsoft around the LXD RESTful API, as a cross-platform container management layer.
Ben Armstrong, a Principal Program Manager Lead at Microsoft on the core virtualization and container technologies, has this to say:
“As Microsoft is working to bring Windows Server Containers to the world – we are excited to see all the innovation happening across the industry, and have been collaborating with many projects to encourage and foster this environment. Canonical’s LXD project is providing a new way for people to look at and interact with container technologies. Utilizing ‘system containers’ to bring the advantages of container technology to the core of your cloud infrastructure is a great concept. We are looking forward to seeing the results of our engagement with Canonical in this space.”
Finally, if you're in Seattle next week, we hope you'll join us for the technical sessions we're leading at ContainerCon 2015, including: "Putting the D in LXD: Migration of Linux Containers", "Container Security - Past, Present, and Future", and "Large Scale Container Management with LXD and OpenStack". Details are below.
Date: Wednesday, August 19 10:25am-11:15am Title: Putting the D in LXD: Migration of Linux Containers Speaker: Tycho Andersen Abstract: http://sched.co/3YTz Location: Willow A Schedule: http://sched.co/3YTz
If you read my last post, perhaps you followed the embedded instructions and ran hundreds of LXD system containers on your own Ubuntu machine.
Or perhaps you're already a Docker enthusiast and your super savvy microservice architecture orchestrates dozens of applications among a pile of process containers.
Either way, the massive multiplication of containers everywhere introduces an interesting networking problem:
"How do thousands of containers interact with thousands of other containers efficiently over a network? What if every one of those containers could just route to one another?"
Canonical is pleased to introduce today an innovative solution that addresses this problem in perhaps the most elegant and efficient manner to date! We call it "The Fan" -- an extension of the network tunnel driver in the Linux kernel. The fan was conceived by Mark Shuttleworth and John Meinel, and implemented by Jay Vosburgh and Andy Whitcroft.
A Basic Overview
Each container host has a "fan bridge" that enables all of its containers to deterministically map network traffic to any other container on the fan network. I say "deterministically", in that there are no distributed databases, no consensus protocols, and no more overhead than IP-IP tunneling. [A more detailed technical description can be found here.] Quite simply, a /16 network gets mapped on onto an unused /8 network, and container traffic is routed by the host via an IP tunnel.
A Demo
Interested yet? Let's take it for a test drive in AWS...
First, launch two instances in EC2 (or your favorite cloud) in the same VPC. Ben Howard has created special test images for AWS and GCE, which include a modified Linux kernel, a modified iproute2 package, a new fanctl package, and Docker installed by default. You can find the right AMIs here.
Now, let's create a fan bridge on each of those two instances. We can create it on the command line using the new fanctl command, or we can put it in /etc/network/interfaces.d/eth0.cfg.
We'll do the latter, so that the configuration is persistent across boots.
$ cat /etc/network/interfaces.d/eth0.cfg
# The primary network interface
auto eth0
iface eth0 inet dhcp
up fanctl up 250.0.0.0/8 eth0/16 dhcp
down fanctl down 250.0.0.0/8 eth0/16
$ sudo ifup --force eth0
Now, let's send some traffic back and forth! Again, we can use ping and nc.
root@261ae39d90db:/# ping -c 3 250.0.27.3
PING 250.0.27.3 (250.0.27.3) 56(84) bytes of data.
64 bytes from 250.0.27.3: icmp_seq=1 ttl=62 time=0.563 ms
64 bytes from 250.0.27.3: icmp_seq=2 ttl=62 time=0.278 ms
64 bytes from 250.0.27.3: icmp_seq=3 ttl=62 time=0.260 ms
--- 250.0.27.3 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 1998ms
rtt min/avg/max/mdev = 0.260/0.367/0.563/0.138 ms
root@261ae39d90db:/# echo "here come the bits" | nc 250.0.27.3 9876
root@261ae39d90db:/#
─────────────────────────────────────────────────────────────────────
root@ddd943163843:/# ping -c 3 250.0.28.3
PING 250.0.28.3 (250.0.28.3) 56(84) bytes of data.
64 bytes from 250.0.28.3: icmp_seq=1 ttl=62 time=0.434 ms
64 bytes from 250.0.28.3: icmp_seq=2 ttl=62 time=0.258 ms
64 bytes from 250.0.28.3: icmp_seq=3 ttl=62 time=0.269 ms
--- 250.0.28.3 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 1998ms
rtt min/avg/max/mdev = 0.258/0.320/0.434/0.081 ms
root@ddd943163843:/# nc -l 9876
here come the bits
Alright, so now let's really bake your noodle...
That 250.0.0.0/8 network can actually be any /8 network. It could be a 10.* network or any other /8 that you choose. I've chosen to use something in the reserved Class E range, 240.* - 255.* so as not to conflict with any other routable network.
Finally, let's test the performance a bit using iperf and Amazon's 10gpbs instances!
So I fired up two c4.8xlarge instances, and configured the fan bridge there.
Multiple containers, on separate hosts, directly addressable to one another with nothing more than a single network device on each host. Deterministic routes. Blazing fast speeds. No distributed databases. No consensus protocols. Not an SDN. This is just amazing!
RFC
Give it a try and let us know what you think! We'd love to get your feedback and use cases as we work the kernel and userspace changes upstream.
Over the next few weeks, you'll see the fan patches landing in Wily, and backported to Trusty and Vivid. We are also drafting an RFC, as we think that other operating systems and the container world and the Internet at large would benefit from Fan Networking.
652 Linux containers running on a Laptop? Are you kidding me???
A couple of weeks ago, at the OpenStack Summit in Vancouver, Canonical released the results of some scalability testing of Linux containers (LXC) managed by LXD.
Ryan Harper and James Page presented their results -- some 536 Linux containers on a very modest little Intel server (16GB of RAM), versus 37 KVM virtual machines.
Ryan has published the code he used for the benchmarking, and I've used to to reproduce the test on my dev laptop (Thinkpad x230, 16GB of RAM, Intel i7-3520M).
I managed to pack a whopping 652 Ubuntu 14.04 LTS (Trusty) containers on my Ubuntu 15.04 (Vivid) laptop!
The system load peaked at 1056 (!!!), but I was using merely 56% of 15.4GB of system memory. Amazingly, my Unity desktop and Byobu command line were still perfectly responsive, as were the containers that I ssh'd into. (Aside: makes me wonder if the Linux system load average is accounting for container process correctly...)
As for KVM, I managed to launch 31 virtual machines without KSM enabled, and 65 virtual machines with KSM enabled and working hard. So that puts somewhere between 10x - 21x as many containers as virtual machines on the same laptop.
You can now repeat these tests, if you like. Please share your results with #LXD on Google+ or Twitter!
I'd love to see someone try this in AWS, anywhere from an m3.small to an r3.8xlarge, and share your results ;-)
Density test instructions
## Install lxd
$ sudo add-apt-repository ppa:ubuntu-lxc/lxd-git-master
$ sudo apt-get update
$ sudo apt-get install -y lxd bzr
$ cd /tmp
## At this point, it's a good idea to logout/login or reboot
## for your new group permissions to get applied
## Grab the tests, disable the tools download
$ bzr branch lp:~raharper/+junk/density-check
$ cd density-check
$ mkdir lxd_tools
## Periodically squeeze your cache
$ sudo bash -x -c 'while true; do sleep 30; \
echo 3 | sudo tee /proc/sys/vm/drop_caches; \
free; done' &
## Run the LXD test
$ ./density-check-lxd --limit=mem:512m --load=idle release=trusty arch=amd64
## Run the KVM test
$ ./density-check-kvm --limit=mem:512m --load=idle release=trusty arch=amd64
As for the speed-of-launch test, I'll cover that in a follow-up post!
Our Canonical colleague Stephane Graber posted a bit more technical design detail here on the lxc-devel mailing list, which was picked up by HackerNews. And LWN published a story yesterday covering another Canonical colleague of ours, Serge Hallyn, and his work on Cgroups and CGManager, all of which feeds into LXD. As it happens, Stephane and Serge are upstream co-maintainers of Linux Containers. Tycho Andersen, another colleague of ours, has been working on CRIU, which was the heart of his amazing demo this week, live migrating a container running the cult classic 1st person shooter, Doom! between two containers, back and forth.
Moreover, we've answered a few journalists' questions for excellent articles on ZDnet and SynergyMX. Predictably, El Reg is skeptical (which isn't necessarily a bad thing). But unfortunately, The Var Guy doesn't quite understand the technology (and unfortunately uses this article to conflate LXD with other random Canonical/Ubuntu complaints).
In any case, here's a bit more about LXD, in my own words...
Our primary design goal with LXD, is to extend containers into process based systems that behave like virtual machines.
We love KVM for its total machine abstraction, as a full virtualization hypervisor. Moreover, we love what Docker does for application level development, confinement, packaging, and distribution.
But as an operating system and Linux distribution, our customers are, in fact, asking us for complete operating systems that boot and function within a Linux Container's execution space, natively.
Linux Containers are essential to our reference architecture of OpenStack, where we co-locate multiple services on each host. Nearly every host is a Nova compute node, as well as a Ceph storage node, and also run a couple of units of "OpenStack overhead", such as MySQL, RabbitMQ, MongoDB, etc. Rather than running each of those services all on the same physical system, we actually put each of them in their own container, with their own IP address, namespace, cgroup, etc. This gives us tremendous flexibility, in the orchestration of those services. We're able to move (migrate, even live migrate) those services from one host to another. With that, it becomes possible to "evacuate" a given host, by moving each contained set of services elsewhere, perhaps a larger or smaller system, and then shut down the unit (perhaps to replace a hard drive or memory, or repurpose it entirely).
Containers also enable us to similarly confine services on virtual machines themselves! Let that sink in for a second... A contained workload is able, then, to move from one virtual machine to another, to a bare metal system. Even from one public cloud provider, to another public or private cloud!
The last two paragraphs capture a few best practices that what we've learned over the last few years implementing OpenStack for some of the largest telcos and financial services companies in the world. What we're hearing from Internet service and cloud providers is not too dissimilar... These customers have their own customers who want cloud instances that perform at bare metal equivalence. They also want to maximize the utilization of their server hardware, sometimes by more densely packing workloads on given systems.
As such, LXD is then a convergence of several different customer requirements, and our experience deploying some massively complex, scalable workloads (a la OpenStack, Hadoop, and others) in enterprises.
The rapid evolution of a few key technologies under and around LXC have recently made this dream possible. Namely: User namespaces, Cgroups, SECCOMP, AppArmor, CRIU, as well as the library abstraction that our external tools use to manage these containers as systems.
LXD is a new "hypervisor" in that it provides (REST) APIs that can manage Linux Containers. This is a step function beyond where we've been to date: able to start and stop containers with local commands and, to a limited extent, libvirt, but not much more. "Booting" a system, in a container, running an init system, bringing up network devices (without nasty hacks in the container's root filesystem), etc. was challenging, but we've worked our way all of these, and Ubuntu boots unmodified in Linux Containers today.
Moreover, LXD is a whole new semantic for turning any machine -- Intel, AMD, ARM, POWER, physical, or even a virtual machine (e.g. your cloud instances) -- into a system that can host and manage and start and stop and import and export and migrate multiple collections of services bundled within containers.
I've received a number of questions about the "hardware assisted" containerization slide in my deck. We're under confidentiality agreements with vendors as to the details and timelines for these features.
What (I think) I can say, is that there are hardware vendors who are rapidly extending some of the key features that have made cloud computing and virtualization practical, toward the exciting new world of Linux Containers. Perhaps you might read a bit about CPU VT extensions, No Execute Bits, and similar hardware security technologies. Use your imagination a bit, and you can probably converge on a few key concepts that will significantly extend the usefulness of Linux Containers.
As soon as such hardware technology is enabled in Linux, you have our commitment that Ubuntu will bring those features to end users faster than anyone else!
If you want to play with it today, you can certainly see the primitives within Ubuntu's LXC. Launch Ubuntu containers within LXC and you'll start to get the general, low level idea. If you want to view it from one layer above, give our new nova-compute-flex (flex was the code name, before it was released as LXD), a try. It's publicly available as a tech preview in Ubuntu OpenStack Juno (authored by Chuck Short, Scott Moser, and James Page). Here, you can launch OpenStack instances as LXC containers (rather than KVM virtual machines), as "general purpose" system instances.
Finally, perhaps lost in all of the activity here, is a couple of things we're doing different for the LXD project. We at Canonical have taken our share of criticism over the years about choice of code hosting (our own Bazaar and Launchpad.net), our preferred free software licence (GPLv3/AGPLv3), and our contributor license agreement (Canonical CLA). [For the record: I love bzr/Launchpad, prefer GPL/AGPL, and am mostly ambivalent on the CLA; but I won't argue those points here.]
These have been very deliberate, conscious decisions, lobbied for and won by our engineers leading the project, in the interest of collaborating and garnering the participation of communities that have traditionally shunned Canonical-led projects, raising the above objections. I, for one, am eager to see contribution and collaboration that too often, we don't see.
Docker 1.0.1 is available for testing, in Ubuntu 14.04 LTS!
Docker 1.0.1 has landed in the trusty-proposed archive, which we hope to SRU to trusty-updates very soon. We would love to have your testing feedback, to ensure both upgrades from Docker 0.9.1, as well as new installs of Docker 1.0.1 behave well, and are of the highest quality you have come to expect from Ubuntu's LTS (Long Term Stable) releases! Please file any bugs or issues here.
Moreover, this new version of the Docker package now installs the Docker binary to /usr/bin/docker, rather than /usr/bin/docker.io in previous versions. This should help Ubuntu's Docker package more closely match the wealth of documentation and examples available from our friends upstream.
Here are a few commands that might help your testing...
Check What Candidate Versions are Available
$ sudo apt-get update
$ apt-cache show docker.io | grep ^Version:
If that shows 0.9.1~dfsg1-2 (as it should), then you need to enable the trusty-proposed pocket.
$ echo "deb http://archive.ubuntu.com/ubuntu/ trusty-proposed universe" | sudo tee -a /etc/apt/sources.list
$ sudo apt-get update
$ apt-cache show docker.io | grep ^Version:
And now you should see the new version, 1.0.1~dfsg1-0ubuntu1~ubuntu0.14.04.1, available (probably in addition to 1.0.1~dfsg1-0ubuntu1~ubuntu0.14.04.1).
Upgrades
Check if you already have Docker installed, using:
If you're already a Docker user, you probably don't need these instructions. But in case you're reading this, and trying Docker for the first time, here's the briefest of quick start guides :-)
That's how we roll our operating systems in this modern, bountiful era of broadly deployed virtual machines, densely packed with system containers.
Linux, and more generally free software, is a natural fit in this model where massive scale is the norm. And particularly Ubuntu (with its solid Debian base), is perfectly suited to this brave new world.
Introduced in Ubuntu 8.04 LTS (Hardy) -- November 19, 2007, in fact -- JeOS (pronounced, "juice") was the first of its kind. An absolutely bare minimal variant of the Ubuntu Server, tailored to perfection for virtual machines and appliances. Just enough OS.
Taken aback, I overheard a technical executive at a Fortune 50 company say this week:
"What ever happened to that Ubuntu JeOS thing? We keep looking at CoreOS and Atomic, but what we really want is just a bare minimal Ubuntu server."
Somehow, somewhere along the line, an important message a got lost. I hope we can correct that now...
JeOS has been here all along, in fact. You've been able to deploy a daily, minimal Ubuntu image, all day, every single day for most of the the last decade. Sure, it changed names to Ubuntu Core along the way, but it's still the same sleek little beloved ubuntu-minimal distribution.
That's 63 MB, including a package management system, with one-line, apt-get access to over 30,000 freely available packages across the Ubuntu universe.
That's pretty darn small. Much smaller than say, 165 MB or 268 MB (which, to be fair, includes a bit more of an operating system -- much closer to say the standard Ubuntu Cloud Image, which is a 176 MB root tarball, or with kernel at 243 MB).
"How useful could such a small image actually be, in practice?", you might ask...
Oh, and guess what else? Ubuntu Core is available for more than just the amd64 architecture! It's also available for i386, armhf, arm64, powerpc, and ppc64el. Which is pretty cool, particularly for embedded systems.
So next time you're looking for just enough operating system, just look to the core. Ubuntu Core. There is truly no better starting point ;-)
Dustin Kirkland (Twitter, LinkedIn) is an engineer at heart, with a penchant for reducing complexity and solving problems at the cross-sections of technology, business, and people.
With a degree in computer engineering from Texas A&M University (2001), his full-time career began as a software engineer at IBM in the Linux Technology Center working on the Linux kernel and security certifications, including a one-year stint as an dedicated engineer-in-residence at Red Hat in Boston (2005). Dustin was awarded the title Master Inventor at IBM, in recognition of his prolific patent work as an inventor and reviewer with IBM's intellectual property attorneys.
Dustin then first joined Canonical (2008) as an engineer (eventually, engineering manager), helping create the Ubuntu Server distribution and establishing Ubuntu as the overwhelming favorite Linux distribution in Amazon, Google, and Microsoft's cloud platforms, as well as authoring and maintaining dozens of new open source packages.
Dustin joined Gazzang (2011), a venture-backed start-up built around an open source project that he co-authored (eCryptFS), as Chief Technology Officer, and helped dozens of enterprise customers encrypt their data at rest and securely manage their keys. Gazzang was acquired by Cloudera (2014).
Having effectively monetized eCryptFS as an open source project at Gazzang, Dustin returned to Canonical (2013) as the VP of Product for Ubuntu and spent the next several years launching a portfolio of products and services (Ubuntu Advantage, Extended Security Maintenance, Canonical Livepatch, MAAS, OpenStack, Kubernetes) that continues to deliver considerable annual recurring revenue. With Canonical based in London, an 800+ work-from-home employee roster and customers spread across 40+ countries, Dustin traveled the world over, connecting with clients and colleagues steeped in rich cultural experiences.
Google Cloud (2018) recruited Dustin from Canonical to product manage Google's entrance into on-premises data centers with its GKE On-Prem (now, Anthos) offering, with a specific focus on the underlying operating system, hypervisor, and container security. This work afforded Dustin a view deep into the back end data center of many financial services companies, where he still sees tremendous opportunities for improvements in security, efficiencies, cost-reduction, and disruptive new technology adoption.
Seeking a growth-mode opportunity in the fintech sector, Dustin joined Apex Clearing (now, Apex Fintech Solutions) as the Chief Product Officer (2019), where he led several organizations including product management, field engineering, data science, and business partnerships. He drastically revamped Apex's product portfolio and product management processes, retooling away from a legacy "clearing house and custodian", and into a "software-as-a-service fintech" offering instant brokerage account opening, real-time fractional stock trading, a secure closed-network crypto solution, and led the acquisition and integration of Silver's tax and cost basis solution.
Drawn back into a large cap, Dustin joined Goldman Sachs (2021) as a Managing Director and Head of Platform Product Management, within the Consumer banking division, which included Marcus, and the Apple and GM credit cards. He built a cross-functional product management community and established numerous documented product management best practices, processes, and anti-patterns.
Dustin lives in Austin, Texas, with his wife Kim and their wonderful two daughters.
