Conversation
667a411 to
cd4a2d0
Compare
cd4a2d0 to
ff82e26
Compare
cmd/dockerd/daemon_unix.go
Outdated
There was a problem hiding this comment.
Maybe planned clean-up, but extra Sprintf
daemon/daemon_unix.go
Outdated
There was a problem hiding this comment.
How differs from DefaultRuntimeBinary?
There was a problem hiding this comment.
This is the name used for the --runtime option map
daemon/info_unix.go
Outdated
There was a problem hiding this comment.
Can we let docker-containerd --version print JSON?
There was a problem hiding this comment.
That's a request for https://github.com/containerd/containerd ;-)
There was a problem hiding this comment.
namespace: docker -> moby? moby-core?
There was a problem hiding this comment.
@mlaventure maybe put "docker" in a variable so we can change it easily.
There was a problem hiding this comment.
Will do, will call the namespace moby too while at it.
libcontainerd/remote_daemon.go
Outdated
There was a problem hiding this comment.
containerd -> value of binaryName?
There was a problem hiding this comment.
If using an already running containerd, that value might be wrong
ff82e26 to
177235a
Compare
pkg/signal/trap.go
Outdated
There was a problem hiding this comment.
Not really, in general we want both together
There was a problem hiding this comment.
This seems like the wrong place to implement this logic. pkg/signal shouldn't really know about the binary names.
I think Trap() needs to be refactored to accept a function to call when it receives SIGQUIT.
There was a problem hiding this comment.
Oh, completely forgot it was inside pkg/trap. Will remove it for now.
a4b39a9 to
263d181
Compare
Those errors are fixed when upgrading the kernel pass 3.13, could we upgrade CI to something a bit more recent? The lts kernel in trusty is 4.4.0 now (cc @andrewhsu) |
|
all 💚 😇 |
cpuguy83
left a comment
There was a problem hiding this comment.
LGTM
I am worried about the upgrade case as the new shim protocol is completely different, and runtime state is also in a completely different location there doesn't appear to be a way to successfully upgrade without killing all running containers, which is problematic for live-restore.
Right now it looks like we basically just end up abandoning the pre-1.0 shims and the container processes end up blocking on I/O once the kernel buffer limit is hit.
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
|
@cpuguy83 it's effectively the best solution:
Maybe the install script could take care of this, it should at least be clearly stated in the release note when it comes to that. |
|
Should we merge this asap or wait for 1.0 GA? |
vdemeester
left a comment
There was a problem hiding this comment.
LGTM 🦁 but same question as @AkihiroSuda
daemon/daemon.go
Outdated
daemon/daemon.go
Outdated
There was a problem hiding this comment.
While doing the modification so I could remember what was the context 😅. I can remove it.
|
@AkihiroSuda Would like to get it in ASAP (i.e. today). Since this is only using the runtime portion the risk is low except for the incompatibility with the previous version. |
cpuguy83
left a comment
There was a problem hiding this comment.
I'm ok that an upgrade breaks people who have live-restore on the previous version, while this has generally worked in the past it really should never be a guarantee.... people consuming moby will need to make sure to handle this properly.
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
Signed-off-by: Kenfe-Mickael Laventure <mickael.laventure@gmail.com>
|
It seems that there are enough LGTMs. Move it to |
|
ping @crosbymichael |
|
I think there are several breaking back compatibility. |
|
@allencloud changing the runtime binary is still supported by containerd 1.0 You will notice that I did not remove any test, and setting the runtime binary is part of them. I've replicated all of the previous features, there should be no regressions normally. If there is, it means we're lacking tests for it. |
|
LGTM |
|
Wait; why did Janky show it passed? |
|
@thaJeztah from the jenkins-job config, not sure why this is the case: |
|
The
|
|
Oh, possibly because those tests are only used in the internal e2e suite, yes. I got confused, was looking why the vendor check didn't fail on this PR (https://github.com/moby/moby/pull/35283/files#r146650223) edit: see @dnephin's comment, our comments just crossed |
- What I did
Replaced containerd v0.2.x with v1.0
- How I did it
Refactored libcontainerd
- How to verify it
CI must be green
- Description for the changelog
Move to containerd v1.0
- A picture of a cute animal (not mandatory but encouraged)
🐼
--
What's not finished
Closes #34662