-
Notifications
You must be signed in to change notification settings - Fork 109
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
Improve status spec #15
Conversation
@khaosdoctor let me know when you're ready for a review on this. I'll hold off for now since it's a draft. |
@arschles Just testing it now, but the code is ready. Will update the status here. What I did was to create a condition:= *v1alpha1.CreateCondition(v1alpha1.Error,v1.ConditionFalse,v1alpha1.AppServiceTerminationError).SetMessage(err.Error())
httpso.AddCondition(condition).SaveStatus(logger, rec.Client) |
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
* logging Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * Adding helm delete functionality to makefile Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * Splitting up functionality, logs and more Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * checking error fetching services Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * Using the proper service name Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * enabling RBAC on endpoints so the scaler can access interceptor queue sizes Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * logs and TODOs Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * using IPs instead of hostnames in external scaler Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * backing off polling interval Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * logging on external scaler startup Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * removing superfluous logging in scaler Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * port and updating namespaces Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * the right namespace Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * adding grpc reflection Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * Makefile targets for creating/deleting example HTTPScaledObject Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * Removing namespace from example HTTPScaledObject file it is specified when you add it from the make target Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * adding keda creation logic to Makefile (and install doc) Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * helm delete keda target Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net> * Passing queue pinger around as a pointer Otherwise the queue lengths are always reported as 0 Signed-off-by: Aaron Schlesinger <aaron@ecomaz.net>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Updated all commits due to DCO check, I think it needs to be re-run |
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
@zroubalik @arschles Regarding the conditions, I think we should follow a different approach on this, please tell me what you think of it. Instead of using conditions and status as we're doing now (almost as a log) I suggest we use the Kubernetes Event API to post events as logs and have only two conditions:
In the future, we can use the status field to add the ingress host and/or service IP for the user app. All other logs such as: Creating Deployment, Creating Scaler and so on would be logged as events using the |
@khaosdoctor bear in mind, that Events are removed from k8s etcd after some period of time and checking operator log could be pretty anoying as the log could be quite huge :) On the other hand checking Status on some resource is quite easy and the information there is pernament 🤷♂️ |
Seeing from this perspective you're right, let's stick with the statuses 😅 |
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
As @coderanger mentioned at the #keda-dev channel, everytime we update the status on the resource it is rescheduled to be reconciled again and it gets us in an infinite loop unless we use predicates, which needed controller version to be at least on 0.6.4 Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Signed-off-by: Lucas Santos <lhs.santoss@gmail.com>
Hey everyone. I'll close this PR because I'll open it in another branch so it does not make the development of new features blocked by it. |
chore(deps): update dependency fluxcd/flux2 to v0.39.0
Fixes #9