It makes the deployment fast if the images are pre-cached on the instance.
I am trying to make the linkerd development quicker on my VMs
A command similar to this
kubeadm config images pull --kubernetes-version 1.18.8
Something like this maybe?
linkerd images pull --version 2.8.1
What do you want to happen? Add any considered drawbacks.
A for loop :)
for i in web controller proxy grafana; do
docker pull gcr.io/linkerd-io/${i}:stable-2.8.1
done
docker pull gcr.io/linkerd-io/proxy-init:1.3.3
Hi @vikas027 the bin directory includes several scripts to support developing and building Linkerd. I like the idea of using your for loop workaround as a script in the bin directory because, it's simple and effective.
Since the use case for this feature is specifically for development, I don't think the linkerd CLI is the right tool to pull images.
We'd love a PR, if you're interested in submitting one 馃槃
Hey @cpretzer,
Since the use case for this feature is specifically for development
The use case is not for a development environment. I am one of the team members who manage hundreds of AWS accounts for different teams and build AMIs for them. The teams do not have access to pull images from docker.io, so we have to give them AMIs pre-cached with docker images, the process of which is obviously automated.
Now, with every Linkerd release, I will have to find out all the dependent docker images manually and then cache them into the AMI. I know the effort is not huge but good to have everything automated :)
Can I find all the images needed by a Linkerd release programmatically? If there is a way, my task is done.
Thanks,
Vikas
@vikas027 thanks for clarifying the use case. I only saw the development use case in the OP 馃槃
I have just what you're looking for to get the images programmatically:
linkerd install --ignore-cluster | grep image: | sed -e 's/^ *//' | sort | uniq
Credit goes to @wmorgan for that one 馃挜
Thanks, @cpretzer for the hint. This fulfils my need :)
~ # linkerd install --ignore-cluster | grep image: | sed -e 's/^ *//' | sort | uniq | awk -F': ' '{ print $2 }' | sed '/^$/d' | xargs -L 1 docker pull
~ #
Most helpful comment
Thanks, @cpretzer for the hint. This fulfils my need :)