On M1 Mac, to start an existing container, I always have to call start twice. The first time I get an error, and the second time always succeeds. This is for any container. Here is the output from starting

swaggergen

and

plantuml

> docker start swaggergen
FATA[0000] failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: time="2022-10-03T14:35:34Z" level=fatal msg="failed to call cni.Setup: plugin type=\"bridge\" failed (add): failed to allocate for range 0: 10.4.0.223 has been allocated to default-5bc005115dca3ab9857e16c6c3dbf3f219269ea65628c7b974f0d98f1c7b5755, duplicate allocation is not allowed"
Failed to write to log, write /var/lib/nerdctl/dbb19c5e/containers/default/5bc005115dca3ab9857e16c6c3dbf3f219269ea65628c7b974f0d98f1c7b5755/oci-hook.createRuntime.log: file already closed: unknown
> docker start swaggergen
swaggergen

> docker start plantuml
FATA[0000] failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: time="2022-10-03T14:39:17Z" level=fatal msg="failed to call cni.Setup: plugin type=\"bridge\" failed (add): failed to allocate for range 0: 10.4.0.43 has been allocated to default-92f1c588f142c0f2457ad20b4aef575227a8c3ee85d9d53f1aef2f31ba158826, duplicate allocation is not allowed"
Failed to write to log, write /var/lib/nerdctl/dbb19c5e/containers/default/92f1c588f142c0f2457ad20b4aef575227a8c3ee85d9d53f1aef2f31ba158826/oci-hook.createRuntime.log: file already closed: unknown
> docker start plantuml
plantuml

Hi, are you still having this problem? If so, could you create an issue on github? That way we can get the detail needed to debug the problem

It appears to have gone away with 1.6.0 or 1. Btw, when I migrated to 1.6.1 I couldn't restart my sqledge container and had to rebuild it. It would get a similar error but a second try didn't help. You can see I have the container, then try to start it. (Mac M1)

nerdctl ps -a
CONTAINER ID    IMAGE                                       COMMAND                   CREATED           STATUS     PORTS                                             NAMES
cae064719108    <http://mcr.microsoft.com/azure-sql-edge:latest|mcr.microsoft.com/azure-sql-edge:latest>     "/opt/mssql/bin/perm…"    3 months ago      Created    0.0.0.0:1433->1433/tcp                            sqledge
nerdctl start sqledge
FATA[0000] failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error mounting "/var/lib/nerdctl/dbb19c5e/volumes/default/01d2561c214100328fe36a76c5727495a70cb504014d75c59fb1e67f5972df46/_data" to rootfs at "/var/opt/mssql-extensibility/data": stat /var/lib/nerdctl/dbb19c5e/volumes/default/01d2561c214100328fe36a76c5727495a70cb504014d75c59fb1e67f5972df46/_data: no such file or directory: unknown

It appears to have gone away with 1.6.0 or 1.

That's great 😄

Btw, when I migrated to 1.6.1 I couldn't restart my sqledge container and had to rebuild it.

That's odd. Is this still a problem? Can you reproduce it? If the answer to either of those questions is yes, it would be great if you could create an issue

Since I re-created the containers, I haven't seen it again. If I get time I can try downgrading to test the scenario, but I haven't heard others mention it.

Ok, well let us know if you see it again!

It's baaaack. I restarted Rancher and when starting stopped containers I had to do it two or three times.

> docker start sqledge
time="2022-11-02T12:49:48Z" level=fatal msg="cannot access containerd socket \"/run/k3s/containerd/containerd.sock\": no such file or directory"
FATA[0000] cannot access containerd socket "/run/k3s/containerd/containerd.sock": no such file or directory
> docker start sqledge
FATA[0000] failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: time="2022-11-02T12:50:50Z" level=fatal msg="failed to call cni.Setup: plugin type=\"bridge\" failed (add): failed to allocate for range 0: 10.4.0.56 has been allocated to default-ca265bf4ef4cac3f93938ac4def55b8f238ce1ba0cc8eb961d086c081ad678ac, duplicate allocation is not allowed"
Failed to write to log, write /var/lib/nerdctl/dbb19c5e/containers/default/ca265bf4ef4cac3f93938ac4def55b8f238ce1ba0cc8eb961d086c081ad678ac/oci-hook.createRuntime.log: file already closed: unknown
> docker start sqledge
sqledge

> docker start azuresqledge
FATA[0000] failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: time="2022-11-02T12:51:41Z" level=fatal msg="failed to call cni.Setup: plugin type=\"bridge\" failed (add): failed to allocate for range 0: 10.4.0.55 has been allocated to default-cea2e8e137a0d2039055b00814174843c3cc260f90417dffd98aa2a2d9faa255, duplicate allocation is not allowed"
Failed to write to log, write /var/lib/nerdctl/dbb19c5e/containers/default/cea2e8e137a0d2039055b00814174843c3cc260f90417dffd98aa2a2d9faa255/oci-hook.createRuntime.log: file already closed: unknown
> docker start azuresqledge
azuresqledge

failed to allocate for range 0: 10.4.0.56 has been allocated to default-ca265bf4ef4cac3f93938ac4def55b8f238ce1ba0cc8eb961d086c081ad678ac, duplicate allocation is not allowed

Sounds like something isn't getting deleted. Either that, or you have something else running that is trying to use this address range? That doesn't seem likely though. Are you running Docker Desktop, or any other container software?

Hmm, no, doesn't sound familiar to me…

@helpful-butcher-75556 does this issue occur on 1.5.1? I know it's intermittent, but if you don't need the features in 1.6.1 you could try running 1.5.1 for a while. Might help us narrow down what is causing the problem

I'm pretty sure it was on 1.5.1 since I thought it went away with 1.6.0

Hmm it doesn't seem likely that that would turn up any new info then. I think it would be best to just create an issue in github. I'm stumped

Ok, thanks. I'll add an issue to RD and try to watch for a consistent way to reproduce it