cardano-api: 9.1 -> 9.3 #1686

locallycompact · 2024-10-07T10:09:25Z

CHANGELOG updated or not needed
Documentation updated or not needed
Haddocks updated or not needed
No new TODOs introduced or explained herafter

We intend to create one of the hydra-plutus scripts with aiken.

This separates the aiken build, which yields the plutus.json blueprint and the haskell build which provides access to the compiled validators.

Instead of using the plutus-tx "compiledScript", we use the loaded-from-blueprint aiken script in the init validator and when constructing/observing transactions in the hydra-node.

This is not yet working and it does not correctly detect changes to plutus.json

Plutus-tx seems to be encoding tuples as a Constr on-chain.

Also add TODO comments for things that are missing to be removed.

… script info

By regenerating the plutus.json file using -k flag during aiken build.

This removes the need of using aiken/builtin function decode_utf8, providing some small optimization. Also recreated plutus scripts.

To correlate with the rest.

Also removed its golden spec. Finally added a note about a fragility being present while accessing the commit validator script from the blueprintJSON.

Seems like the aiken compiler uses plutus core version 1.1.0 syntax and we need to declare it as a PlutusV3 script.

We do not need the plutus-style envelope and this would not match with the PlutusV2 used there.

This will keep the first trace argument which we expect to see in our mutation tests.

This does not require us to be in a git repository and hence makes the test more flexible and isolated.

The hydra-plutus library seems to be consistently rebuilt when the plutus.json gets changed.

This was failing with PPViewHashesDontMatch because the transaction was referencing the commit script which is does not always need, and the commit script being of different version than the other scripts spent.

github-actions · 2024-10-07T10:27:10Z

Transaction costs

Sizes and execution budgets for Hydra protocol transactions. Note that unlisted parameters are currently using arbitrary values and results are not fully deterministic and comparable to previous runs.

Metadata
Generated at	2024-10-07 10:24:05.783152986 UTC
Max. memory units	14000000
Max. CPU units	10000000000
Max. tx size (kB)	16384

Script summary

Name	Hash	Size (Bytes)
νInitial	b512161ccb0652d7e9a0b540e4a3c808f73d6558a4bcabf374d85880	3969
νCommit	ea444d37d226e71eef73ac78d149750da977feb588900135bf9e8221	692
νHead	2253ddd95837c7aacc8635a971caaea743434152dd8dd2849bdf4162	10797
μHead	4d648ca239040b0e87901835aa11423e7aa3bd947ce6befe7db1bae8*	4508
νDeposit	1a011f23b139a6426767026bde10319546485d553219a5848cdac4e5	2993

The minting policy hash is only usable for comparison. As the script is parameterized, the actual script is unique per head.

`Init` transaction costs

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	5097	5.81	2.30	0.44
2	5295	6.99	2.76	0.46
3	5502	8.37	3.30	0.48
5	5899	11.41	4.51	0.53
10	6906	18.11	7.16	0.65
57	16356	82.89	32.79	1.78

`Commit` transaction costs

This uses ada-only outputs for better comparability.

UTxO	Tx size	% max Mem	% max CPU	Min fee ₳
1	569	10.84	4.26	0.29
2	754	14.31	5.80	0.34
3	946	17.92	7.39	0.39
5	1320	25.56	10.73	0.49
10	2249	47.11	19.97	0.77
19	3937	94.71	39.81	1.38

`CollectCom` transaction costs

Parties	UTxO (bytes)	Tx size	% max Mem	% max CPU	Min fee ₳
1	56	559	19.84	7.58	0.39
2	114	671	27.94	10.64	0.48
3	171	782	34.42	13.13	0.56
4	226	893	42.47	16.18	0.65
5	283	1004	48.66	18.55	0.72
6	337	1116	61.67	23.44	0.87
7	394	1227	73.57	27.91	1.00
8	450	1338	86.71	32.85	1.15
9	507	1449	88.96	33.80	1.18

Cost of Decrement Transaction

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	625	18.64	8.15	0.39
2	807	20.66	9.71	0.42
3	896	20.63	10.42	0.43
5	1124	22.64	12.78	0.47
10	2051	33.45	20.79	0.66
49	7765	95.49	75.11	1.80

`Close` transaction costs

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	672	20.83	9.32	0.42
2	741	21.93	10.44	0.44
3	911	23.70	12.18	0.47
5	1278	27.30	15.69	0.54
10	2090	35.41	23.62	0.70
50	7927	96.94	84.74	1.89

`Contest` transaction costs

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	683	26.77	11.46	0.48
2	798	28.61	13.05	0.51
3	994	30.76	14.93	0.55
5	1225	33.78	17.67	0.60
10	1960	43.29	26.04	0.77
38	6511	97.56	74.18	1.76

`Abort` transaction costs

There is some variation due to the random mixture of initial and already committed outputs.

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	4997	15.51	6.64	0.54
2	5089	21.89	9.41	0.62
3	5334	29.67	12.98	0.72
4	5348	34.39	14.90	0.77
5	5583	42.07	18.42	0.87
6	5649	47.61	20.71	0.93
7	5664	50.15	21.55	0.96
8	5861	61.01	26.45	1.09
9	6094	67.42	29.38	1.17
10	6014	70.86	30.55	1.21
11	6373	85.53	37.25	1.39
12	6265	85.92	37.04	1.39
13	6533	90.60	39.31	1.45

`FanOut` transaction costs

Involves spending head output and burning head tokens. Uses ada-only UTxO for better comparability.

Parties	UTxO	UTxO (bytes)	Tx size	% max Mem	% max CPU	Min fee ₳
10	0	0	5089	9.99	4.18	0.49
10	1	56	5122	11.35	4.98	0.50
10	5	284	5258	15.63	7.71	0.56
10	10	569	5429	21.28	11.24	0.64
10	20	1138	5768	33.55	18.72	0.81
10	30	1709	6111	45.04	25.88	0.97
10	40	2275	6446	57.33	33.37	1.14
10	50	2844	6786	68.44	40.36	1.29
10	77	4380	7700	99.74	59.82	1.73

End-to-end benchmark results

This page is intended to collect the latest end-to-end benchmark results produced by Hydra's continuous integration (CI) system from the latest master code.

Please note that these results are approximate as they are currently produced from limited cloud VMs and not controlled hardware. Rather than focusing on the absolute results, the emphasis should be on relative results, such as how the timings for a scenario evolve as the code changes.

Generated at 2024-10-07 10:26:56.564616191 UTC

Baseline Scenario

Number of nodes	1
Number of txs	3000
Avg. Confirmation Time (ms)	4.699312577
P99	8.134082609999997ms
P95	5.9774916ms
P50	4.2668275ms
Number of Invalid txs	0

Three local nodes

Number of nodes	3
Number of txs	9000
Avg. Confirmation Time (ms)	22.620816611
P99	104.98001816000011ms
P95	28.687728549999978ms
P50	20.500917ms
Number of Invalid txs	0

github-actions · 2024-10-07T10:31:31Z

Test Results

518 tests ±0 512 ✅ ±0 29m 44s ⏱️ + 6m 26s
163 suites ±0 6 💤 ±0
7 files ±0 0 ❌ ±0

Results for commit e659c67. ± Comparison against base commit d3db623.

This pull request removes 1 and adds 1 tests. Note that renamed tests count towards both.

Hydra.Options/Hydra Node RunOptions ‑ validateRunOptions: using more than 5 parties should error out

Hydra.Options/Hydra Node RunOptions ‑ validateRunOptions: using more than 8 parties should error out

locallycompact and others added 30 commits October 4, 2024 18:26

Update cardano-api: 9.1 -> 9.3

3e288c3

Add aiken to our nix develop shell

bbee1ba

We intend to create one of the hydra-plutus scripts with aiken.

Draft an aiken project in hydra-plutus

c9e19c9

Draft an empty commit aiken validator and commit plutus.json

4343b66

Create a module to load blueprint json

3b98df5

This separates the aiken build, which yields the plutus.json blueprint and the haskell build which provides access to the compiled validators.

Replace plutus-tx commit with the aiken commit script

9f4ac0f

Instead of using the plutus-tx "compiledScript", we use the loaded-from-blueprint aiken script in the init validator and when constructing/observing transactions in the hydra-node.

implement 'ViaCollectCom' path in Aiken

ef8b7d5

Implement 'ViaAbort' road on commit validator in Aiken.

4c6bc03

Try to make plutus.json a extra source file

eb368a7

This is not yet working and it does not correctly detect changes to plutus.json

Decode commit datum as a record

f706f7a

Plutus-tx seems to be encoding tuples as a Constr on-chain.

Add golden spec to verify plutus blueprint

0e320b1

Add commit errors to the validator script

f0551ca

Add comments to the aiken commit validator script

7e3e680

Remove commit validator

e7fb038

Also add TODO comments for things that are missing to be removed.

Add test to check plutus blueprint hash against the code reference in…

15f587d

… script info

Alter plutus.json and aiken.lock

b2eef70

Fix commit mutation tests on abort and collect-com

0607633

By regenerating the plutus.json file using -k flag during aiken build.

Remove unit-test focus

0585f6e

Replace the use of ByteArray for String on aiken commit validator script

69547b6

This removes the need of using aiken/builtin function decode_utf8, providing some small optimization. Also recreated plutus scripts.

Increase the number of parties for FanOut tx-cost

895905b

To correlate with the rest.

Increase the maximum number of parties to 7

4a22a58

Remove commit plutus validator as it is not needed anymore

244f815

Also removed its golden spec. Finally added a note about a fragility being present while accessing the commit validator script from the blueprintJSON.

Upgrade aiken project version to match next release

54f1a26

Add a CIP-57 todo

210f65c

Re-add golden tests for commit validator

a3f4865

Add aiken as a build dep to hydra-plutus tests

4ad41dc

Update changelog

7ff4ecc

Compiling aiken script

5c4c79f

Pretty print reference scripts in renderTxWithUTxO

e4851b2

Use aiken scripts as PlutusV3 script

2527fac

Seems like the aiken compiler uses plutus core version 1.1.0 syntax and we need to declare it as a PlutusV3 script.

ch1bo and others added 11 commits October 4, 2024 18:26

Add/update plutus v3 cost models

8d5590e

Drop commit from inspect-script exe

d49a2a2

We do not need the plutus-style envelope and this would not match with the PlutusV2 used there.

Rebuild commit aiken validator with -t compact

c52c7af

This will keep the first trace argument which we expect to see in our mutation tests.

Simply compare the plutus.json after regeneration

90b32b8

This does not require us to be in a git repository and hence makes the test more flexible and isolated.

Drop a redundant test and FIXME

55dd765

The hydra-plutus library seems to be consistently rebuilt when the plutus.json gets changed.

Fix publishHydraScripts to use V3 for the commit script

db10f25

Only reference commit script if spending commits in abort

75a0775

This was failing with PPViewHashesDontMatch because the transaction was referencing the commit script which is does not always need, and the commit script being of different version than the other scripts spent.

Update changelog and set maximumNumberOfParties to 9

f4d379a

Bump versions and regenerate scripts

f9e07ce

Only 8 participants are possible consistently

665cc17

cardano-api: 9.1 -> 9.3

e659c67

locallycompact force-pushed the lc/cardano-api-9.3 branch from deb0481 to e659c67 Compare October 7, 2024 10:20

locallycompact closed this Oct 7, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

cardano-api: 9.1 -> 9.3 #1686

cardano-api: 9.1 -> 9.3 #1686

locallycompact commented Oct 7, 2024

github-actions bot commented Oct 7, 2024

github-actions bot commented Oct 7, 2024

cardano-api: 9.1 -> 9.3 #1686

cardano-api: 9.1 -> 9.3 #1686

Conversation

locallycompact commented Oct 7, 2024

github-actions bot commented Oct 7, 2024

Transaction costs

Script summary

Init transaction costs

Commit transaction costs

CollectCom transaction costs

Cost of Decrement Transaction

Close transaction costs

Contest transaction costs

Abort transaction costs

FanOut transaction costs

End-to-end benchmark results

Baseline Scenario

Three local nodes

github-actions bot commented Oct 7, 2024

Test Results

`Init` transaction costs

`Commit` transaction costs

`CollectCom` transaction costs

`Close` transaction costs

`Contest` transaction costs

`Abort` transaction costs

`FanOut` transaction costs