ExecutionEffort and Computation is not the same in the stable cadence emulator

[bjartek]

Instructions

Please fill out the template below to the best of your ability and include a label indicating which tool/service you were working with when you encountered the problem.

Problem

Feed decuted event looks like this

Fields: map[string]interface {}{
        "amount": 1e-05,
        "executionEffort": 2.9e-07,
        "inclusionEffort": 1.0,
      },
      

so 29 effort

Meters reported looks like this

map[string]interface {}{
  "computationIntensities": map[string]interface {}{
    "1001": 5.0,
    "1003": 5.0,
    "1025": 2.0,
    "1026": 96.0,
    "1080": 250.0,
    "2005": 1.0,
    "2011": 1.0,
    "2014": 1.0,
    "2015": 1.0,
    "2016": 1.0,
    "2020": 2286.0,
    "2022": 1.0,
    "2026": 270.0,
    "2034": 3.0,
    "2036": 1.0,
  },
  "computationUsed": 31.0,
  "ledgerInteractionUsed": 97269.0,
  "memoryEstimate": 149745.0,
  "memoryIntensities": map[string]interface {}{
    "1": 7.0,
    "100": 6.0,
    "102": 5.0,
    "103": 694.0,
    "105": 376.0,
    "106": 2.0,
    "107": 244.0,
    "108": 30.0,
    "109": 13.0,
    "11": 1.0,
    "110": 8.0,
    "111": 11.0,
    "112": 8.0,
    "115": 20.0,
    "116": 24.0,
    "117": 27.0,
    "119": 8.0,
    "12": 2.0,
    "120": 4.0,
    "127": 9.0,
    "128": 1.0,
    "129": 1.0,
    "13": 1.0,
    "130": 8.0,
    "131": 4.0,
    "133": 12.0,
    "137": 5.0,
    "138": 1.0,
    "139": 4.0,
    "14": 3.0,
    "140": 4.0,
    "147": 4.0,
    "148": 7.0,
    "149": 1.0,
    "15": 8.0,
    "151": 2.0,
    "153": 108.0,
    "154": 23.0,
    "155": 38.0,
    "156": 2.0,
    "158": 1.0,
    "159": 19.0,
    "16": 5.0,
    "162": 1.0,
    "165": 3.0,
    "166": 1.0,
    "17": 1.0,
    "172": 29.0,
    "173": 1.0,
    "174": 2.0,
    "176": 8.0,
    "177": 2236.0,
    "178": 1176.0,
    "179": 2.0,
    "18": 8.0,
    "180": 12.0,
    "181": 8.0,
    "182": 9.0,
    "187": 5.0,
    "19": 5.0,
    "23": 5.0,
    "24": 0.0,
    "25": 161.0,
    "30": 2.0,
    "31": 2.0,
    "32": 8.0,
    "34": 2.0,
    "39": 3.0,
    "4": 88.0,
    "41": 8.0,
    "5": 6.0,
    "7": 5.0,
    "8": 4.0,
    "9": 1.0,
    "94": 1034.0,
    "95": 5.0,
    "96": 284.0,
    "97": 11.0,
    "98": 2.0,
  },
}

so this reports 31

These should be the same should the not?

Steps to Reproduce

Run a transaction against the emulator with the relevant version
github.com/onflow/flow-emulator v0.59.1-0.20231218185945-9116c416533f

Acceptance Criteria

Effort and computation are the same.

Context

I am trying to update overflow to this code and I see that the emulator is reporting different numbers for FeesDeducted and computationUsed, is that exepected? because it has not been like that before this versio

[bjartek]

I see the same in 0.59.0 of emulator

[janezpodhostnik]

I'm trying t understand this better and I'm replicating the scenario, but I'm missing something.

• I got the emulator at commit 9116c41
• run it with go run ./cmd/emulator --transaction-fees=true --verbose
• send a tx with flow transactions send <(echo """transaction{prepare(s : &Account){ var i = 0;while i < 10000 {i=i+1}}}""") --signer emulator-account
• the resulting event has executionEffort (UFix64): 0.00000478
• which matches the emulator output:
  DBG ⭐ Transaction executed computationUsed=478 memoryEstimate=25369033 txID=b5b11e587630d03db246596381739b35cafff3a147e84fb6497e397b6f96cd0a

But I'm not sure at which point you are capturing the meter output. If you are capturing the meter output to late the system transaction might be included in it.

p.s.: I tried the fix from @bluesign #543 but that doesn't change anything in my scenario

[janezpodhostnik]

never mind, figured out whats going on after I manually enabled logging the computation intesities. Yeah #543 will fix this.