[FVM] Add normalised time per computation logs to transaction execution #5385
Conversation
| // log the normalized time per computation | ||
| // if the computation estimation is correct the value should be 1. | ||
| // if the value is greater than 1, the computation estimation is too low; we are underestimating transaction complexity (and thus undercharging). | ||
| // if the value is less than 1, the computation estimation is too high; we are overestimating transaction complexity (and thus overcharging). |
There was a problem hiding this comment.
Does it matter if the transaction failed? Which might have a much smaller time per computation?
Meaning we might not be overcharging, but simply because the transaction failed.
There was a problem hiding this comment.
we should be charging correctly even if the transaction fails. The tx still takes time before failing.
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #5385 +/- ##
==========================================
+ Coverage 55.98% 56.03% +0.04%
==========================================
Files 1026 1018 -8
Lines 99902 99473 -429
==========================================
- Hits 55930 55737 -193
+ Misses 39677 39446 -231
+ Partials 4295 4290 -5
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
|
are there some example outliers? |
|
@bluesign here are some. https://www.flowdiver.io/tx/0f66b1a024d11a3fb44c1eb2c4c7f61fa1ff5cdea4d828f5301e39c561229a2b?tab=script |
The normalised time per computation metrics that were added are proving to be useful (https://flowfoundation.grafana.net/goto/a_9XFKTSg?orgId=1), but it would be great to have these logs, so it would be easier to chase down outliers.