Backport e2e-test of the latest tendermint main branch

[tnasu]

Description

e2e-test is old and it makes too many failures each PR and Push. I investigated the error logs of e2e-test, and it shows light01 failure due to starting delay. light01 depends on the witness("validator01", "validator02", "validator03") and validator02 depends on the seeds("seed02"). And the log shows that "seed02" is in the process of starting normally. Therefore, "validator02" and "light01" were also starting up and did not reach the expected Height within the expected time, resulting in an error. So "seed02" is removed.

• light01 happened an error

light01        | I[2022-12-21|02:55:43.981] Downloading trusted light block using options module=main 
light01        | I[2022-[12](https://github.com/line/ostracon/actions/runs/3738979308/jobs/6360920253#step:10:13)-21|02:55:43.985] error comparing first header with witness. You may want to consider removing the witness module=main err="light block not found"
light01        | panic: runtime error: invalid memory address or nil pointer dereference
light01        | [signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0xc220b4]
light01        | 
light01        | goroutine 30 [running]:
light01        | github.com/line/ostracon/light/provider/http.(*http).LightBlock(0xc000ba1a40, {0x153f440, 0xc000b64c40}, 0x0)
light01        | 	/src/ostracon/light/provider/http/http.go:86 +0x[13](https://github.com/line/ostracon/actions/runs/3738979308/jobs/6360920253#step:10:14)4
light01        | github.com/line/ostracon/light.(*Client).getTargetBlockOrLatest(0xc000ba1a40?, {0x153f440, 0xc000b64c40}, 0x3e8, {0x153c3a0, 0xc000ba1a40})
light01        | 	/src/ostracon/light/detector.go:382 +0x3c
light01        | github.com/line/ostracon/light.(*Client).compareNewHeaderWithWitness(0xc00003e840, {0x153f440, 0xc000b64c40}, 0xc00005ff78?, 0xc000537a70, {0x153c3a0, 0xc000ba1a40}, 0x0)
light01        | 	/src/ostracon/light/detector.go:136 +0x285
light01        | created by github.com/line/ostracon/light.(*Client).compareFirstHeaderWithWitnesses
light01        | 	/src/ostracon/light/client.go:1[14](https://github.com/line/ostracon/actions/runs/3738979308/jobs/6360920253#step:10:15)5 +0x96b

• validator02 is starting now, it's not started, so light01 cannot get a block

validator02    | I[2022-12-21|02:55:26.107] received complete proposal block             module=consensus height=1000 hash=158E96A848B4AF973BAFAF7453B3077DEDE33044D49C327B44A853A101C86E88
validator02    | I[2022-12-21|02:55:29.105] Timed out                                    module=consensus dur=3s height=1000 round=0 step=RoundStepPropose
validator02    | I[2022-12-21|02:55:50.053] Starting Peer service                        module=p2p peer=2551a13ed720101b271a5df4816d1e4b3d3bd133@[fd80:b10c::6]:52926 impl="Peer{MConn{[fd80:b10c::6]:52926} 2551a13ed720101b271a5df4816d1e4b3d3bd133 in}"
validator02    | I[2022-12-21|02:55:50.053] Starting MConnection service                 module=p2p 

• seed02 is started at "02:55:50", so validator02 cannot start

seed02         | I[2022-12-21|02:55:50.602] commit is for a block we do not know about; set ProposalBlock=nil module=consensus height=1000 commit_round=1 proposal= commit=B0278D2A75529747D37EFB72D40A9A30AB6EEC716BAE3C490F3DB45F9246428B
seed02         | I[2022-12-21|02:55:50.703] received complete proposal block             module=consensus height=1000 hash=B0278D2A75529747D37EFB72D40A9A30AB6EEC716BAE3C490F3DB45F9246428B
seed02         | I[2022-12-21|02:55:50.704] finalizing commit of block                   module=consensus height=1000 hash=B0278D2A75529747D37EFB72D40A9A30AB6EEC716BAE3C490F3DB45F9246428B 

Based on: tendermint/tendermint@b1fbe26

• e2e-test: https://github.com/tendermint/tendermint/tree/b1fbe26ffadae6e4010a71bd75767c6680d1bf71/test/e2e
• e2e-test history: https://github.com/tendermint/tendermint/commits/b1fbe26ffadae6e4010a71bd75767c6680d1bf71/test/e2e

Refs:

• https://github.com/tendermint/tendermint/pull/9292/files#r950183182

Remarks: (Excludes some of the new features)

• feat: v0.34.x Prioritized Mempool tendermint/tendermint#8695
• abci: implement process proposal to spec tendermint/tendermint#9122
• remove v1 and v2 blocksync protocol impementations tendermint/tendermint#9146
• config: backport the rename of fastsync to blocksync  tendermint/tendermint#9259
• proto: deduplicate consensus params tendermint/tendermint#9287
• abci: Port EventAttribute field type change to main tendermint/tendermint#9335
• inspect: resurrect the inspect command tendermint/tendermint#9655
• abci: Adapt unsynchronized local client to replicate remote client concurrency tendermint/tendermint#9830

[codecov]

Codecov Report

Merging #536 (8b6b7a6) into main (67311dc) will increase coverage by 0.00%.
The diff coverage is 92.00%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #536   +/-   ##
=======================================
  Coverage   66.27%   66.28%           
=======================================
  Files         280      281    +1     
  Lines       38133    38147   +14     
=======================================
+ Hits        25273    25285   +12     
+ Misses      11063    11048   -15     
- Partials     1797     1814   +17     

Impacted Files	Coverage Δ
test/e2e/generator/main.go	55.00% <70.00%> (+55.00%)	⬆️
test/e2e/generator/random.go	98.18% <92.85%> (ø)
libs/log/lazy.go	100.00% <100.00%> (ø)
libs/log/ocfmt_logger.go	84.28% <100.00%> (+0.70%)	⬆️
test/e2e/generator/generate.go	90.41% <100.00%> (-0.76%)	⬇️
libs/events/events.go	93.23% <0.00%> (-5.27%)	⬇️
privval/socket_listeners.go	79.72% <0.00%> (-4.06%)	⬇️
blockchain/v0/pool.go	79.06% <0.00%> (-3.11%)	⬇️
privval/signer_listener_endpoint.go	88.88% <0.00%> (-2.39%)	⬇️
... and 12 more

[Kynea0b]

Simple question. Please tell me what you think. @tnasu
Q1. About fixing timing errors
Since validator02 depends on seed02, light01, which depends on validator02, cannot get a block if seed02 is delayed. So, validator02 is set to depend on seed01 like validator03 and validator01.

Is it a change that lowered the probability of failure due to delay with the above change?

Q2. About post-commit errors
https://github.com/line/ostracon/actions/runs/3749344661/jobs/6367896870#step:9:12

E[2022-12-21|12:55:35.620] timed out waiting for seed01 to reach height 0

Looking at this error, does this mean that if seed01 is late it can error out as well?

[tnasu]

@kokeshiM0chi

Q1

Yes, it does. Because we cannot control the network condition on GitHub Actions.

Q2

Yes, it does. But, if seed01 happens an error, it returns an error immediately and we can re-run e2e-test immediately. light02's error is so late and waits a long time.

[ulbqb]

LGTM

[Kynea0b]

LGTM

[torao]

Tip: The test/maverick using the Byzantine node has been removed in this PR. In subsequent modifications, it's not necessary to also modify the maverick package when modifying consensus or node modules.