V1 Deprecation: Remove algod and kmd client affinity and consolidate APIs to use newest version

[algochoi]

Summary

As part of sunsetting V1 algod APIs, we want goal to only use V2 algod APIs. This PR removes client affinity and consolidates all algod API calls to use the V2 version (mostly in test calls). kmd affinity is also removed, but this does not affect any functionality since kmd only has one version.

This does not change any goal behavior, as it fetched the V2 algod client by default: https://github.com/algorand/go-algorand/pull/4641/files#diff-75c865ad4f92ef3366534a2489ccd8aaf03e31a1b17027d64fe2216076d3c067L384

Test Plan

CI Tests.

[codecov]

Codecov Report

Merging #4641 (c7f6cea) into master (bb7c59f) will increase coverage by 0.04%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #4641      +/-   ##
==========================================
+ Coverage   54.34%   54.39%   +0.04%     
==========================================
  Files         402      402              
  Lines       51793    51786       -7     
==========================================
+ Hits        28149    28168      +19     
+ Misses      21276    21253      -23     
+ Partials     2368     2365       -3     

Impacted Files	Coverage Δ
cmd/goal/commands.go	10.44% <ø> (+0.03%)	⬆️
libgoal/libgoal.go	2.70% <ø> (+0.02%)	⬆️
data/transactions/verify/txn.go	76.19% <0.00%> (-0.96%)	⬇️
network/wsNetwork.go	64.76% <0.00%> (+0.19%)	⬆️
catchup/service.go	69.62% <0.00%> (+0.74%)	⬆️
agreement/proposalManager.go	98.03% <0.00%> (+1.96%)	⬆️
agreement/cryptoVerifier.go	69.71% <0.00%> (+2.11%)	⬆️
network/wsPeer.go	67.64% <0.00%> (+2.13%)	⬆️
ledger/roundlru.go	96.22% <0.00%> (+5.66%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[jannotti]

I would approve except for the question of users of MakeRestClient - we need to make sure, for example, that we aren't making the loadgenerator slower.

[jannotti]

Do we understand why this was using APIVersionV1? It means, for example, that the loadgenerator cmd/loadgenerator/main.go uses V1 today.

[michaeldiamant]

I'm not sure.

• Scanning git history, I see 1750fb5#diff-c8e8e0d825d7542580daf188e1c5f6b4f079c888d10dfb2544a8a75b6419d0b4R72. When version affinity was introduced, MakeRestClient was locked to v1.
• My unsubstantiated guess is: Using v1 simplified other ongoing work at the time and switching to v2 got lost in the shuffle.
• Additionally - For load generator algod API usage, I scanned endpoint implementations in v1 + v2 and spotted no meaningful differences. Of course, it'd be preferable for someone more familiar with load generator to vouch and/or run a test.

[algorandskiy]

@cce @brianolson could you look into it to double check the loadgenerator does not break?

[brianolson]

pingpong has several V1 dependencies

[brianolson]

I didn't do a performance stress test but loadgenerator does basically work with this change in.

loadgenerator -d node -config '{"RoundModulator":10,"Fee":10000,"TxnsToSend":10}'

[algochoi]

@brianolson Thank you for checking - visually, it doesn't look like the v2 API will cause any harm. Are reviewers OK with moving forward with this change then?

[michaeldiamant]

Barring objections, I'll merge the PR near EOD.