Well, I imagine given the current love of Docker on this site you're going to be down-voted with a quickness, but from what I've seen and done with it so far I really don't see the great benefit of it over standard LXC. Essentially it's a neat wrapper and has some cool API features, but I don't see the benefit beyond that.
Honestly if I was/am going to hitch my cart to LXC I'd probably be going the libvirt-lxc route so I'd be dealing with something of an actual standard.
Edit: I suppose one reason people might prefer Docker over libvirt would be no XML blegh, but it's honestly not that bad.
It seems like the core of Docker is little more than a helper script that should be part of libvirt or LXC proper (which btw, already has a system for building containers from a template), but that would involve less horn-blowing and web design than it would submitting boring, well-tested old patches upstream.
DotCloud have done a wonderful job of popularizing the idea of LXC containers with Docker. I think the idea of LXC containers as software-distribution mechanism is a great one, and I credit Docker with getting that idea into my mind.
It is great to see that many of Docker's ideas are being implemented upstream; e.g. LXC added support for BtrFS snapshots shortly after Docker launched. Sounds like OpenShift is considering which Docker ideas belong in their platform.
It can sometimes be hard to work with upstream projects, who often have a different world-view. I do like the idea of releasing a "sacrificial lamb" project that is a demonstration of your ideas, even if long-term all the ideas belong upstream (i.e. LXC in the case of Docker).
I think jQuery is a good analogy. One day jquery might be unnecessary, when all browsers everywhere implement all the high-level API goodness. In the meantime... :)
Here's my honest experience. Half a year ago, I used to think like that too. Then I tried everything vanilla. I used naked lxc. I used those lxc standard commands like lxc-start. I dedicated a partition as btrfs for lxc because I needed a lot of nodes based on the same base image. Then I wrote scripts to automate everything. It had been a fun experience. But it just stopped there. None of what I did was general enough to be reused. Then 3 months later, I tried docker again. I started to appreciate it. It achieve the same but was with much less effort. Documentation is not the best, but the community is really responsive. I dumped my own scripts and never looked at them again.
Comments
Well, I imagine given the current love of Docker on this site you're going to be down-voted with a quickness, but from what I've seen and done with it so far I really don't see the great benefit of it over standard LXC. Essentially it's a neat wrapper and has some cool API features, but I don't see the benefit beyond that.
Honestly if I was/am going to hitch my cart to LXC I'd probably be going the libvirt-lxc route so I'd be dealing with something of an actual standard.
Edit: I suppose one reason people might prefer Docker over libvirt would be no XML blegh, but it's honestly not that bad.
It seems like the core of Docker is little more than a helper script that should be part of libvirt or LXC proper (which btw, already has a system for building containers from a template), but that would involve less horn-blowing and web design than it would submitting boring, well-tested old patches upstream.
DotCloud have done a wonderful job of popularizing the idea of LXC containers with Docker. I think the idea of LXC containers as software-distribution mechanism is a great one, and I credit Docker with getting that idea into my mind.
It is great to see that many of Docker's ideas are being implemented upstream; e.g. LXC added support for BtrFS snapshots shortly after Docker launched. Sounds like OpenShift is considering which Docker ideas belong in their platform.
It can sometimes be hard to work with upstream projects, who often have a different world-view. I do like the idea of releasing a "sacrificial lamb" project that is a demonstration of your ideas, even if long-term all the ideas belong upstream (i.e. LXC in the case of Docker).
I think jQuery is a good analogy. One day jquery might be unnecessary, when all browsers everywhere implement all the high-level API goodness. In the meantime... :)
A great analogy. Hopefully - just like jQuery - you can get Microsoft on board with LXC as well :-)
We're on HackerNews. "it's honestly not that bad"... the point of hacking is taking things that are bad or "not so bad" and making them better.
This argument to me achieves only the opposite, it makes me never want to consider whatever option it defends.
Here's what it looks like to create a container in libvirt-lxc: http://libvirt.org/drvlxc.html
It is pretty dang clear to me how to create it.
omg does this work even with Ruby and Go!?! :O Do you have a tutorial???! The fonts suck so bad. Omg this is not even web scale
Why yes, it actually does have Ruby bindings:
http://libvirt.org/ruby/
Here's my honest experience. Half a year ago, I used to think like that too. Then I tried everything vanilla. I used naked lxc. I used those lxc standard commands like lxc-start. I dedicated a partition as btrfs for lxc because I needed a lot of nodes based on the same base image. Then I wrote scripts to automate everything. It had been a fun experience. But it just stopped there. None of what I did was general enough to be reused. Then 3 months later, I tried docker again. I started to appreciate it. It achieve the same but was with much less effort. Documentation is not the best, but the community is really responsive. I dumped my own scripts and never looked at them again.
Have they solved the issue of non-deterministic commands e.g. apt-get update, which could mean a Docker container behaving differently.