Open olafmol opened 7 years ago
This is related to metrics/health workflows you want to use - e.g. instead of using Vamp default ones (simple count of total and 500 responses), you can customise/refine algorithm. Health ideally should not be based (just) on 500 response codes as this is an extreme. Metrics workflow should filter out some requests etc. This can be a good example for a docs/blog.
f.e. when running Vamp on AWS with ELBs, we configure the ELB to do a health check on the VGA's. This shows up in the metrics and graphs, but needs to be filtered out to give the real "visitors"/usage metrics.