This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.
You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

HTTP polling - gRPC and HTTP/2

Greetings.

My company plans to roll out Red Hat Openshift with associated monitoring components Prometheus and Thanos. 

We plan to continue using Solarwinds as the primary monitoring and alerting tool. 

In the future we want to look at AppOptics for the fullest integration with Prometheus/Thanos. For now we plan to use http script monitors to pull back selected alerts and metrics from that platform.

Thanos documentation says its required to configure the 'ingress controller' to support gRPC and HTTP/2.

Question: will the requirement for gRPC and HTTP/2 affect Solarwinds HTTP polling in any way?  Does Solarwinds support gRPC and HTTP/2?

Thanks in advance for any inputs from the community.

Parents
  • So the ingress controller is not exactly relevant to SW, that's a part of the cluster building thats going to happen where your thanos lives (im kind of assuming its a k8s deployment somewhere), essentially the ingress controller is their load balancer.  Since http2 is backward compatible and mostly the same as http1.1 that's not going to be a blocker for you, the changes are all in optimizing how connections are established but if you made a request with 1.1 you'd still get a response, it just does it in the classic serial pattern instead of being able to multiplex many requests and responses at once.

    So focusing on the part that is specific to Orion, I've seen people attempt to use the SAM API monitor to scrape prom data, but the result is the full contents of the /metrics page and I don't think they've added a way for the API component to parse that into separate metrics yet.  So that basically leaves you needing to use a custom script component to do do an invoke-restmethod to get the contents of the page and then break it apart into the various metrics and format those into a response that Orion will like.

    https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.utility/invoke-restmethod?view=powershell-7.3#example-8-send-a-request-using-http-2-0

Reply
  • So the ingress controller is not exactly relevant to SW, that's a part of the cluster building thats going to happen where your thanos lives (im kind of assuming its a k8s deployment somewhere), essentially the ingress controller is their load balancer.  Since http2 is backward compatible and mostly the same as http1.1 that's not going to be a blocker for you, the changes are all in optimizing how connections are established but if you made a request with 1.1 you'd still get a response, it just does it in the classic serial pattern instead of being able to multiplex many requests and responses at once.

    So focusing on the part that is specific to Orion, I've seen people attempt to use the SAM API monitor to scrape prom data, but the result is the full contents of the /metrics page and I don't think they've added a way for the API component to parse that into separate metrics yet.  So that basically leaves you needing to use a custom script component to do do an invoke-restmethod to get the contents of the page and then break it apart into the various metrics and format those into a response that Orion will like.

    https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.utility/invoke-restmethod?view=powershell-7.3#example-8-send-a-request-using-http-2-0

Children