backend: Add websocket multiplexer #2459
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
dosubot
bot
added
the
size:XL
|
Backend Code coverage changed from 61.6% to 61.6%. Change: 0% 😃. Coverage report |
|
Nice, tested it and it seems to work There are a couple of crucial things missing though:
|
|
Does this support webosocket connections other than watching resources, like attaching to the container? I've tried a path like this |
|
Thank you for the feedback @sniok. Awesome suggestions. I did forget to add closing the connection on request. Will add Regarding attaching, I did not try it. Will check why it’s not working. |
|
I think we can delay any work on attaching to the container if there is a problem. Because we mainly want this to help with watching. |
Hrmm. Not sure why this is at 0, since there are tests... lol. |
knrt10
force-pushed
the
websocket-multiplexer
branch
from
cf0f799 to
3971e72
Compare
|
Backend Code coverage changed from 61.5% to 60.8%. Change: -.7% 😞. Coverage report |
|
Thanks for the changes.
Imagine we started watching two /pods from two different clusters (or two different ReplicaSets,etc) and we need to figure out where to put this update that we received from websocket. Currently we just get a Pod without any info where it came from
|
|
@sniok thank you for the detailed message. I have updated the description, screenshot, postman API and code. You will now have detailed information from where request is coming from. |
This adds websocket multiplexer to the backend. Frontend now make a single websocket call to the backend. Once that connection is established with the backend it will send message to the backend with appropriate data. Backend will open multiple websockets and act as a proxy to the frontend. It will make request to k8s server and return data to frontend. This also adds retry logic if the connection is broken between frontend and backend. Fixes: #1802 Signed-off-by: Kautilya Tripathi <[email protected]>
knrt10
force-pushed
the
websocket-multiplexer
branch
from
3971e72 to
bf2ec6d
Compare
dosubot
bot
added
size:XXL
|
Backend Code coverage changed from 61.5% to 61.1%. Change: -.4% 😞. Coverage report |
|
Thanks @knrt10 (It looks like close can be used, and now a clusterID can be used to identify the response... Or is it also important to identify the specific request, rather than just the cluster?) Does it make sense to start implementing the frontend/ integration now? Will you do that? Note I started an issue here about multi cluster e2e tests here: |
|
Thanks for the explanation @knrt10 So the response returns (clusterID, path, query, and userID) as well as the data. This can be used to know from which websocket it comes from. We discussed the possibility that multiple ones could be made with the same of those... in this case it will not work. I don't think we need to cover this now though. It could be solved in the future by adding a separate unique ID for each one. Again, I don't think this is necessary for our use cases... we can use (clusterID, path, query, and userID) as the unique key. |
sniok
approved these changes
Oct 24, 2024
|
looks good, thanks! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This adds websocket multiplexer to the backend. Frontend now need to make a single websocket call to the backend. Once that connection is established with the backend it will send message to the backend with appropriate data. Backend will open multiple websockets and act as a proxy to the frontend. It will make request to k8s server and return data to frontend.
This also adds retry logic if the connection is broken between frontend and backend.
Fixes: #1802
Testing
cd backend && go run ./cmd -dev -enable-dynamic-clusterscd frontend && npm start(we need this for userID of the session).{ "clusterID": "minikube", // update it with the cluster you have running "path": "/api/v1/pods", "query": "watch=true", "userId": "9872e4f36eaadb0e" // go to localstorage and copy `headlamp-userID` and paste here }Connectbuttonkubectl run nginx --image=nginxand try other things like logs and execLogs
{ "clusterID": "minikube", "path": "/api/v1/namespaces/default/pods/nginx/log", "query": "follow=true&tailLines=100", "userId": "9872e4f36eaadb0e" }Exec
{ "clusterID": "minikube", "path": "/api/v1/namespaces/default/pods/nginx/exec", "query": "command=ls&command=-l&stdout=true&stderr=true", "userId": "9872e4f36eaadb0e" }The data is sent as binary and it is encoded to bas64 to parse as JSON response.
Close connection
{ "clusterID": "minikube", "path": "/api/v1/pods", "userId": "9872e4f36eaadb0e", "data": "close" }You can edit
pathas per your needs and check different output.