Closed consideRatio closed 4 years ago
So currently, if I deploy this, the pod names are workshop-prod-jupyterhub-ssh-sftp-server-6f95fbcb8d-wq9df
and workshop-prod-jupyterhub-ssh-ssh-server-58f9d5bb57-dklr6
. I think that's too redundant... I'll fiddle around to see what I can get the defaults to be!
fwiw, I think the kinda nams I'd like are jupyterhub-ssh
and jupyterhub-sftp
...
More importantly for me, the service names are workshop-prod-jupyterhub-ssh-sftp-server
and workshop-prod-jupyterhub-ssh-ssh-server
, and I don't think those should be the default.
I fiddled with this for a while, but couldn't get it to work :( I think my preference is for the default to produce deployments and services named jupyterhub-ssh
and jupyterhub-sftp
, while allowing admins to add release name as prefix. Do you think we can either take that part out of this PR, or try and fix it?
Thank you so much for working on this, @consideRatio! I really, really appreciate the thoroughness and thoughtfulness you put into this.
Here fullnameOverride isn't set. Then, the default behavior of a Helm chart with would be to name the secret resource which is scoped as common to both the k8s services and deployments as ...
workshop-prod-jupyterhub-ssh
But, since we have multiple deploy/service/netpol resources, I have appended "ssh-server" and "sftp-server" to them, which can be found in _helpers.tpl
, so the name of these resources becomes...
workshop-prod-jupyterhub-ssh-ssh-server workshop-prod-jupyterhub-ssh-sftp-server
Assume fullnameOverride is set to yuvi, then the secret will be named
yuvi
And the two sets of deploy/service/netpol resources will be named
yuvi-ssh-server yuvi-sftp-server
I suggest to stick with this configuration option system, but you can absolutely make fullnameOverride
something by default and change the suffix to separate the two sets of components from ssh-server
to just ssh
for example.
@yuvipanda I've set fullnameOverride: jupyterhub
by default now, which makes services be named jupyterhub-ssh
and jupyterhub-sftp
.
I suggest we merge at this point, thank you for helping me debug things remotely!
AMAZING, @consideRatio. I have been far away from actually working on helm charts for a while, so I deeply deeply appreciate all your work here. <3
PR addressing ideas in #17, closing #17.
@yuvipanda I updated issue #17 to describe everything added in this PR, and I feel at this point that I have done what I felt was reasonable fixes to do right away without pre-optimizing something. It is now ready for review, where each commit may be easier to overview now that there are plenty of changes in a single PR. Each commit is probably quite self-contained.
In addition to whats listed in #17
Suggested but not part of PR
helm template
or so.