Fix pingpong teal test#2835
Conversation
…lding per wallet. Only create one wallet per NPN Host to consolidate value for pingpong testing.
egieseke
requested review from
algojack,
bricerisingalgorand and
onetechnical
Codecov Report
@@ Coverage Diff @@
## master #2835 +/- ##
==========================================
- Coverage 47.12% 47.09% -0.03%
==========================================
Files 349 349
Lines 56434 56434
==========================================
- Hits 26593 26580 -13
- Misses 26859 26869 +10
- Partials 2982 2985 +3
Continue to review full report at Codecov.
|
tsachiherman
left a comment
tsachiherman
left a comment
There was a problem hiding this comment.
looks good to me. Please make sure to regenerate all the scenarios with this change and include them in the PR.
|
Changing the total supply can trigger some failures that would not be possible on mainnet. Are we truely ran out of money? |
+1. I'm still not a fan of increasing the amount of supply, as it indicates an unrealistic test (we would never increase mainnet supply). |
algobarb
left a comment
algobarb
left a comment
There was a problem hiding this comment.
Messaged Eric to learn more about the details. The results look correct on Datadog now so I think the last fix to rebalance stake fixed it! Changes look good to me.
Revert total non participating wallets to 400 as before.
Updated scenario1 and scenario3. |
By rebalancing the stake between participating and non-participating nodes, increasing the total stake is not required. Reverted total stake back to 10 billion Algos. |
algobarb
left a comment
algobarb
left a comment
There was a problem hiding this comment.
Looks like Scenario1 results look ok with the rebalanced stake and removed wallets as well. Looks good to me.
@onetechnical Good point. Updated so that the total stake remains at 10 billion. Split stake evenly between participating and nonparticipating nodes, which provides a sufficient amount for the teal pingpong tests. |
| nonParticipatingNodeStake := ratZero | ||
| if npnHosts > 0 { | ||
| wallets = 2 * wallets | ||
| // split participating an non participating stake evenly |
There was a problem hiding this comment.
Have we always been assuming an even split between participating and non participating nodes? I wonder if that should be configurable. It looks like that would be a new feature though so probably out of scope
There was a problem hiding this comment.
It was evenly split before. I will create a ticket to make it configurable.
There was a problem hiding this comment.
Created this ticket to allow configuration of stake ratio: https://github.com/algorand/go-algorand-internal/issues/1566
7 participants
Summary
Updated the netgoal utility to only create one wallet per non-participating node to better concentrate Algos for use by pingpong.
Test Plan
Ran netgoal to update the scenario2 recipe and tested the result with the CICD performance pipeline.