What would you like to be added:
MCM needs to have a way to configure API requests frequency.
Why is this needed:
On failure to create an instance, MCM makes a lot of API calls to the instance provider server, exhausting API request limits. There is already a similar issue, requesting an exponential back-off timer after a failure.
Yet, I believe the ability to change the timer manually or simply increase it during the problem (which can last if, for example, the cloud provider does not have the necessary capacity at the moment) is also an important one.
First we'll be working on #483, to introduce an exponential backoffs
We'll need to collect metrics #211 for different provider API request calls to figure out which ones are the pain points, to better understand the configurability that we would want to expose to the user.
/area quality /kind enhancement /priority 3
What would you like to be added: MCM needs to have a way to configure API requests frequency.
Why is this needed: On failure to create an instance, MCM makes a lot of API calls to the instance provider server, exhausting API request limits. There is already a similar issue, requesting an exponential back-off timer after a failure.
Yet, I believe the ability to change the timer manually or simply increase it during the problem (which can last if, for example, the cloud provider does not have the necessary capacity at the moment) is also an important one.