Adopt new @capjs-telemetry plugin and remove deprecated application logging service, put instead CLS

[alperdedeoglu]

No description provided.

[AjitKP91]

@alperdedeoglu

Some Quick Notes:

• CLS is not available on trial or free tier
• integration with SAP Cloud Identity Services - Identity Authentication SAML 2.0 is a pre-requisite.

[arajsinha]

@AjitKP91

Researched some parameters we could think of adding that could be useful for providers based on some of the resources you provided to me (Azure, AWS, OpenTelemetry) and based on my own research.

Most of the metrics mentioned separately are all there in our Cloud Logging Service anyway, in one UI or the other.

However, something I noticed we don't have (but can do with OpenTelemetry) that both AWS and Azure have mentioned in their docs too, is data metrics per tenant, that will make it easier for the providers to track errors and which tenant it is exactly happening in.

Some points we could look at:

• Request Count per tenant - Identifies tenants that have very high usage or are going through unusual spikes (could be a part of security too)
• Errors per tenant - Showing specific request errors based on tenants use cases. It could also list the top errors of each tenant which could make it easier to handle errors as a reference, and also share this data in a specific dashboard for each tenant to make their lives easier.
• Track resource consumption per tenant - Identify overuse or inefficiencies (we have autoscaler and rate-limiter, but good to track high usage tenants to try to upsell to them)
• Request Success Rate per tenant - Track request success rate per tenant to understand health metrics
• Pre-Post Deployment Metrics - Use Telemetry to compare pre & post deployment metrics per tenant, and even globally for a given period of time to identify if the application in that tenant, or globally is performing better or atleast the same as before.

We can show 2-3 examples of how we are using tenant-id's and correlation-id's to aggregate and create some of these things, and list to providers how these other metrics might be useful, so that they can follow the whole thing and build their own metrics based on top of that..