You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Wish to send frontend telemetry data directly to victoriametrics endpoints from Alloy.
Architecture is as following :
GKE kubernetes cluster with nodejs frontends
grafana deployed in dev-grafana namespace
Faro and Alloy delployed in dev-faro namespace
victoriametrics prometheus scraping metrics from all applications
Only able to define either tempo or mimir endpoints at the moment for real user monitoring which we do not have in our platform as of now. Neither do we have Loki .
Only got one reference of the same in documention viz.,
**Captured frontend measurements are stored in Loki instead of Prometheus to avoid cardinality issues.**
Captured frontend measurements are stored in Loki instead of Prometheus to avoid cardinality issues.
Proposed solution
Please could it be explained more clearly if we only need to pick Faro and Alloy for Real user monitoring then can we use prometheus OR victoriametrics endpoints directly ? If not why and also please do elaborate on cardinality issues if any referring to the documents mentioned above.
The text was updated successfully, but these errors were encountered:
Faro has it's own wire protocol.
Every signal besides traces send from Faro is stored in Loki. Traces are stored in tempo.
Faro provides an otel-transport which translates Faro payload to otel before sending it but with the exception that it doesn't create otel metrics, only otel-logs and traces.
So long story short best is to use Faro together with Alloy because alloy does the translation and sends the data to Loki/Tempo.
Loki has the ability to create metrics from logs (metric queries).
Description
Wish to send frontend telemetry data directly to victoriametrics endpoints from Alloy.
Architecture is as following :
Only got one reference of the same in documention viz.,
faro-web-sdk/demo/docs/instrumentations/measurements.md
Line 14 in 09ca958
Captured frontend measurements are stored in Loki instead of Prometheus to avoid cardinality issues.
Proposed solution
Please could it be explained more clearly if we only need to pick Faro and Alloy for Real user monitoring then can we use prometheus OR victoriametrics endpoints directly ? If not why and also please do elaborate on cardinality issues if any referring to the documents mentioned above.
The text was updated successfully, but these errors were encountered: