Update Ocelot configuration without downtime about ocelot HOT 3 CLOSED

Online updates to ReRoute, needs to have a unique identity that exists in the code ReRouteKey identifies the ReRoute is now used for load balancing, it is a calculated value, I think there should be certain identification value, similar to a database table's primary key, so that we can store ReRoute in database or Consul, it will be great storage across servers and updates In this case the local profile can be a backup of the currently running.

from ocelot.

@geffzhang first iteration I do not want to use any dependencies such as consul or database. I try to take a lean approach to software development and use dependencies only when I have to.

UpstreamPathTemplate and UpstreamHttpMethod must be unique in Ocelot so we do not need a specific key such as GUID.

My plan is to expose API that lets you post entire configuration and reload it. I know that this is a hack but I want to make sure authentication is in place before we start looking at better ways to do this. When I have something basic working I will make it better.

Then I will make cluster of Ocelots that keep configuration up to date across cluster.

Finally I will make Ocelot into service discovery so services can register with Ocelot.

All of these features will be behind interface so a user can swap them out and use a database/consul etc if they want to.

from ocelot.

I am studying this section, where can I find relevant examples?

from ocelot.