We are investigating how to do message logging in an optimal way from our on-premise Apigee Edge, thereby leveraging the Message Logging and/or Service Callout policies.
Our overall setup has a number of constraints:
- There is no internal syslog or other logging service available
- API proxy developers and support people cannot access the servers and thus cannot access the logs on disk
- All outbound traffic must go over forward proxy
Questions:
- Is there a way to make log files on disk accessible? Is it possible to read logfiles with API proxy, e.g. using Node.js? Any samples available?
- When logging with Message Logging policy to a sylog-in-the-cloud, any suggestions how to get through an forward http proxy?
- When using the Service Callout policy to log over HTTP to e.g. Splunk, logging will be synchronous and thus delaying the overall proxy?
- To arrange for logging in the PostClientFlow, any suggestions on how to temporarily keep all log messages somewhere in memory and then write them out in the PostClientFlow?
4 questions at once, looking forward to your feedback and suggestions!