-
Notifications
You must be signed in to change notification settings - Fork 65
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Migrate Farallon Staging hub to this repository. #379
Conversation
Getting closer! Now at:
|
for Makes sense it expires after 18h! |
#380 fixes the timeouts. Turns out if you don't have a directory named https://pilot-hubs.2i2c.org/en/latest/topic/storage-layer.html has more info about how we do storage and why this matters. |
Great! I was going to suggest looking at mount targets and security groups that sometimes generate timeouts at the time to mount EFS. |
@damianavila for the expiry of the generated kubeconfig, I see two options:
Actually, we can probably do both - do (2) first and follow up with (1) |
I've done (2), and opened #381 to track (1) |
It makes sense to me in the current context. |
Not complete yet
- Needs to be nested under singleuser - Explicitly set mem_limit too - otherwise, the default 1G limit of memory stays in, and k8s doesn't allow a limit to be smaller than a guarantee
EFS doesn't let you set ver=4.2
We mount this into all daskhub user accounts regardless - by default they don't do anything
Needs to be automated eventually
With KUBECONFIG=secrets/farallon.yaml kops export kubecfg --admin=730h farallon-2i2c.k8s.local
Referencing base-hub from helm charts templates is basically impossible - go templates can not use '-' in their name! This makes referencing them easier
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK, I think this one is ready!
Thanks for all the review, @damianavila! |
kops
produces kubeconfig files we can check-in to the repo,and use to authenticate to
kops
clusters on AWS.We test this by moving the farallon staging hub to be deployed
from here, for #368
TODO:
kops
doesn't work after about 12h