Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Docker is dead? Podman – an alternative tool?

content.fme.de

Post reply on HN
Print view View original

161–170 of 183 posts

Re: Docker is dead? Podman – an alternative tool?

#161

Post by sam_lowry_ »

slim wrote: post #125
forty wrote: post #106

Earlier quoted context omitted.

By default container images are not signed (there is notary, but it's not commonly used - maybe notary V2 will change that - and I think the signature changes depending on the registry it's hosted on anyway?) which make it inconvenient to mirror. Now, why are we still producing new package formats without mandatory signatures (containers, npm, cargo, etc) is not really clear to me. I guess everyone must think "those…

There's also the conspiracy : "it's the CIA who infiltrated docker to get access on every server in the world" ;)

This is a quite reasonable attack vector for US cyber-offense.

Re: Docker is dead? Podman – an alternative tool?

#162

Post by Bilal_io »

411111111111111 wrote: post #159
lesam wrote: post #123

Earlier quoted context omitted.

Yeah, the main benefit of docker desktop is packaging up a nice Linux VM on Mac/Windows, plus some UI features. There’s a reason ’Docker Desktop for Linux’ came so much later.

Dunno, I think WSL2 is actually better on that front. Just do the super secure "curl get.docker.com|bash" in it and you're golden. Even better UX then docker desktop if you specifically don't want a GUI

I think there are benefits to using Docker on Windows without relying on WSL2, especially when you need to run windows containers that depend on HypervV. I've worked at a few .Net shops that were stuck with .Net Framework with no clear, not easy way to modernize, so perhaps a compromise is containerize the applications into Windows containers.

Re: Docker is dead? Podman – an alternative tool?

#163

Post by jrop »

ncphil wrote: post #145
KronisLV wrote: post #82

Earlier quoted context omitted.

Makes me think of what other laptops (or even just PC hardware in general) have some of the best support on *nix. Personally, in my experience most hardware has been passably okay, at least when compared with trying to run FreeBSD, though some required tweaking (such as touchpads). I think Lenovo ThinkPad laptops have some of the best reputation in that regard, though the newer models can definitely be a bit on the p…

With ThinkPads, just buy a few generations behind refurbished, preferably from a nonprofit recycler. They've always been great for Linux. The T series gives you more upgrade options, but the X models run longer on a charge and are small and light . I'm on my second now, having "upgraded" from a T430 to an X250. But I miss the larger screen and so will probably get a T470 or T480 soon. Running boring plain vanilla Ubu…

I'm running Pop_OS! (after a breakage with Arch) on my T14s and the experience has been really great.

Re: Docker is dead? Podman – an alternative tool?

#164

Post by sigg3 »

vocram wrote: post #108
smarx007 wrote: post #96

How did the intro get so many things wrong?! 1. Mirantis did not acquire Docker Inc., they only bought Docker Enterprise. See https://techcrunch.com/2019/11/13/mirantis-acquires-docker-e... and https://www.docker.com/blog/docker-enterprise-edition/ 2. k8s didn't remove dockershim for political reasons but because containerd was refactored out of Docker long time ago and k8s wanted to get rid of the extra layer. See h…

> To execute the images Podman then uses e.g. the mentioned containerd, Another wrong thing. Podman directly controls the runtime (crun or runC). It does not talk with containerd like Docker.

Right. This is a major selling point of podman in Red Hat material.

Re: Docker is dead? Podman – an alternative tool?

#165

Post by lioeters »

lapser wrote: post #69

This article misses quite a bit. Podman has a full Docker compatible API, so you just have to enable it, and then set the DOCKER_HOST to point to its socket. From there docker compose should work as if you had Docker. Podman also is currently working on "podman machine", which can spin up a Linux VM to run Podman on macOS and Windows. I think it's still in beta or something, but it seems to be working already. There…

About "podman machine":

> Podman also runs on Mac and Windows, where it provides a native podman CLI and embeds a guest Linux system to launch your containers. This guest is referred to as a Podman machine and is managed with the `podman machine` command.

> ..On Mac, each Podman machine is backed by a QEMU based virtual machine.

> ..On Windows, each Podman machine is backed by a virtualized Windows System for Linux (WSLv2) distribution.

https://podman.io/getting-started/installation.html

Re: Docker is dead? Podman – an alternative tool?

#166

Post by deskamess »

astrostl wrote: post #153
Hamcha wrote: post #56

Earlier quoted context omitted.

On the Docker for Windows note, it's not podman but after I had trouble with their "forced upgrade unless enterprise" policy (which made me update to a broken version with a known issue they didn't solve for weeks) I switched to Rancher Desktop and never looked back. You get all the things (docker CLI, docker-compose, kubernetes via k3s) but it's FOSS and it doesn't feel like they're trying to shove a premium plan do…

Loving Rancher Desktop here too, and even w/o making use of k3s at all. Fantastic drop-in replacement for Docker Desktop w/o the garbage.

Interesting/Great that Rancher Desktop uses WSL - not having to launch a Linux VM (albeit transparent) is nice. I still wish MS would have come up with something native.

Re: Docker is dead? Podman – an alternative tool?

#167

Post by namaria »

tasuki wrote: post #43
kristjansson wrote: post #25

Earlier quoted context omitted.

Runs docker in a VM for you, abstracts that away, confuses the hell out of new and old developers alike.

Wasn't the main selling point of docker "no more VMs"?

IMHO doing development work on any long lived environment (especially if also used for personal internet browsing) is a mistake. Embracing disposable VMs for everything has saved me so much time and headache - just scrap it and start anew. Then my personal compute space remains very simple and debris free.

Re: Docker is dead? Podman – an alternative tool?

#168

Post by afiori »

forty wrote: post #106
stingraycharles wrote: post #99

Earlier quoted context omitted.

> The rate limits were mostly in place to deny corporate CI users access to the Hub free of charge and force them to pay or deploy a mirror. What I never understood is why they didn’t just properly handle this with mirrors like any package manager does; why is this a problem for docker, but not for yum / apt / etc? I have to admit that these rate limits have accelerated my migration to alternatives like quay.io

By default container images are not signed (there is notary, but it's not commonly used - maybe notary V2 will change that - and I think the signature changes depending on the registry it's hosted on anyway?) which make it inconvenient to mirror. Now, why are we still producing new package formats without mandatory signatures (containers, npm, cargo, etc) is not really clear to me. I guess everyone must think "those…

I believe that adding key handling as a necessary step for npm, docker, and similar registries isn't a good idea.

But it would be easy to sign all packages: if the author has decided not to use signatures the registry adds it's own signature, at any moment the author can choose to start using their own keys in place of the registrie's

Re: Docker is dead? Podman – an alternative tool?

#169

Post by znpy »

lolcat_cowsay wrote: post #154
znpy wrote: post #55

Earlier quoted context omitted.

You can point the official docker-compose at podman now! I do that! It's 99.999% compatible as the podman people basicaly reimplemented all the docker daemon APIs. It sometimes lags a bit behind, because sometime docker implements new stuff... But for usage with docker-compose it has worked flawlessly for me. EDIT: you can also export the podman unix socket via socat, i also tried it to run a rootless docker runtime…

I tried podman compose quite a few months ago with my docker compose file and it failed, what is the difference between doing that and this? So like, you can use docker compose for podman instead of docker, instead of something like podman compose?

The difference is that instead of using podman-compose you use the actual docker-compose.

You have to point your $DOCKER_HOST to the podman unix socket or something, but other than that it’s the actual docker-compose experience. Via the env var trick you could even use the actual docker binary. But you could just alias docker to podman and it works the same, by design!

Post reply on HN