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
In addition to CPU/memory/network utilization, it will be very useful to show metrics recorded inside each helper. For instance, one can monitor the number of multiplications sitting inside zero-knowledge proofs, infrastructure buffer sizes, number of steps running in parallel, total number of records sent, etc.
This will require draft to talk to helper's backend periodically and collect this information. We may want to leverage some existing infrastructure for it, for example Prometheus. Currently, we don't have any endpoint that can expose these metrics and we don't collect them, so this task may serve as an umbrella to implement metric controls for both IPA and Draft
The text was updated successfully, but these errors were encountered:
In addition to CPU/memory/network utilization, it will be very useful to show metrics recorded inside each helper. For instance, one can monitor the number of multiplications sitting inside zero-knowledge proofs, infrastructure buffer sizes, number of steps running in parallel, total number of records sent, etc.
This will require draft to talk to helper's backend periodically and collect this information. We may want to leverage some existing infrastructure for it, for example
Prometheus
. Currently, we don't have any endpoint that can expose these metrics and we don't collect them, so this task may serve as an umbrella to implement metric controls for both IPA and DraftThe text was updated successfully, but these errors were encountered: