Orchestrator Generator multiple requests from single deploy

[dallas-fletchall]

What product are you having troubles with?

Console

What Console version are you using?

1.119.11

Description

While using the orchestrator generator, we notice that on every deploy we get 2 requests with the same data. The only difference between the requests the we see is the x-request-id.

Using socat, we were able to validate that we are getting multiple requests for each deploy from the front-end.

Actual Outcome

First request (not including payload)

host: REDACTED
user-agent: got (https://github.com/sindresorhus/got)
content-type: application/json
authorization: REDACTED
accept: application/json
content-length: 130044
accept-encoding: deflate,gzip,br
x-forwarded-for: REDACTED
x-forwarded-proto: https
x-request-id: b35214a0-ed01-4e2b-91cd-517faa365978
x-envoy-external-address: REDACTED
x-forwarded-host: REDACTED
scheme: https
x-real-ip: REDACTED
x-original-uri: /helmifier/convert
original-request-uri: /helmifier/convert
original-request-method: POST
miausergroups: unlogged
x-envoy-original-path: /helmifier/convert

Second request

POST /convert HTTP/1.1
host: REDACTED
user-agent: got (https://github.com/sindresorhus/got)
content-type: application/json
authorization: REDACTED
accept: application/json
content-length: 130044
accept-encoding: deflate,gzip,br
x-forwarded-for: REDACTED
x-forwarded-proto: https
x-request-id: 2728c953-869a-422e-acdd-7547f8af59af
x-envoy-external-address: REDACTED
x-forwarded-host: REDACTED
scheme: https
x-real-ip: REDACTED
x-original-uri: /helmifier/convert
original-request-uri: /helmifier/convert
original-request-method: POST
miausergroups: unlogged
x-envoy-original-path: /helmifier/convert

Expected Outcome

A single request for each deploy, or a better understanding of why with updated documentation for orchestrator generator.

[KarissaBrummell]

Hello! We are now using Mia-Platform v14.4.0 and this duplicate call is causing us issues. Looking at logs, we see more than 20 second delays between the 2 calls, and this results in deployments hanging which is an inconvenience for our users. Can someone explain why 2 requests are made and potentially remove the duplicate? Thank you!

[ThisIsDemetrio]

Hi there, @KarissaBrummell .

The double request to the orchestrator unfortunately happens because of the nature of the deployment features in the Console. While the first request is required to actually generate the manifest, the second one is necessary to fetch information to be included in the collection containing the history of previous deployments, to allow to deploy previous saved configurations, which is handled in a separate process (despite being still synchronous - thus the deployment hanging because of the delays).

Of course it is a process that can be improved to reduce the number of requests, despite 20 seconds passing between requests seems a lot. We will analyze the issue and give it back to you once we have a plan and an estimated date on how to improve this process.

Have a nice day,
|D|