Ci experimentation

[pgrange]

This branch introduces an alternative CI process: It's doing the same job as ci.yaml but trying to take benefit of github action
pre-installed software on Ubunut images and cache mechanism.

In my experiments, it's sometimes able to achieve better performance than our current ci, I think it's related to a difference in cache performance between GitHub action and nix.

At this stage, I want us to have both CI live together to observe, after some time, what happens:

• I'm expecting this CI2 to perform better when incrementally commit stuff to a branch
• But it might perform worse when rebasing a branch
• GitHub cache is limited to 10Gb, it could work for us but how much extra build time do we pay if it's not the case?
• Also note that this branch drastically reduces the space we use on GitHub cache so we should see less cache misses here

Both CI and CI2 rely on subtle caching mechanisms and it's only by running them both that we can appreciate the real difference, if any.

These are the changes introduced:

• building the static binary has been extracted in a dedicated workflow, similar to what we do with docker
• a new ci2.yaml defines a similar workflow as ci.yaml but relying on different tooling
• a setup_dev_ubuntu GitHub action has been introduced to limit copy/pasting ci code

So let's merge that, run both of them together for two weeks and get metrics out of it.

---

• No new TODOs introduced or explained herafter

-> TODO review in two weeks after the merge

Oh, I almost forgot, the first build of this on master will take 1h30 or something and this is expected.

[github-actions]

Transactions 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	2023-06-26 09:55:59.730715913 UTC
Max. memory units	14000000
Max. CPU units	10000000000
Max. tx size (kB)	16384

Script summary

Name	Hash	Size (Bytes)
νInitial	2212a4ee618434b9b2f366d7c330dbdfb5c7072e793a850fd0de6ddd	4294
νCommit	69e1ccf9ad73dc6d37a5bc8de5aec86f3c4c1710fe5fd334e0e16b18	2124
νHead	8ae095dca4d14a1b8edffb37faa6c84ec60340fbf389a62f027e0b76	9355
μHead	33642a45c7fbb955ce1704ef09229bb211bf9af9980530db28c6aafe*	4148

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

Cost of Init Transaction

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	4739	12.42	4.85	0.49
2	4946	15.14	5.89	0.53
3	5151	19.61	7.65	0.59
5	5566	24.52	9.51	0.66
10	6584	34.12	13.05	0.81
37	12123	97.21	36.81	1.73

Cost of Commit Transaction

This is using ada-only outputs for better comparability.

UTxO	Tx size	% max Mem	% max CPU	Min fee ₳
1	599	14.89	5.70	0.34
2	787	19.57	7.70	0.40
3	974	24.75	9.88	0.46
5	1349	36.15	14.59	0.61
10	2279	71.73	28.85	1.04
13	2857	98.03	39.15	1.35

Cost of CollectCom Transaction

Parties	UTxO (bytes)	Tx size	% max Mem	% max CPU	Min fee ₳
1	57	814	27.89	10.83	0.49
2	114	1135	43.42	16.99	0.67
3	169	1454	61.77	24.33	0.89
4	225	1775	82.30	32.57	1.13

Cost of Close Transaction

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	635	18.58	8.31	0.39
2	804	19.97	9.56	0.42
3	766	18.75	7.91	0.39
5	1299	24.16	13.33	0.50
10	2124	31.12	19.62	0.64
50	8726	87.17	69.99	1.74

Cost of Contest Transaction

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	667	24.34	10.47	0.45
2	840	26.06	11.84	0.48
3	1005	28.20	13.38	0.52
5	1337	31.63	16.12	0.58
10	2160	39.77	22.80	0.73
45	7926	99.74	70.74	1.81

Cost of Abort Transaction

Some variation because of random mixture of still initial and already committed outputs.

Parties	Tx size	% max Mem	% max CPU	Min fee ₳
1	4930	23.28	9.83	0.63
2	5177	36.59	15.53	0.79
3	5498	53.72	22.99	0.99
4	5818	73.48	31.60	1.23
5	6139	95.31	41.18	1.49

Cost of FanOut Transaction

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 ₳
5	0	0	4765	8.66	3.57	0.46
5	1	57	4796	10.06	4.39	0.47
5	5	283	4943	15.64	7.69	0.55
5	10	567	5121	22.61	11.82	0.64
5	20	1139	5480	36.56	20.07	0.83
5	30	1706	5838	50.52	28.33	1.02
5	40	2275	6202	64.49	36.60	1.21
5	50	2849	6566	78.46	44.87	1.40
5	65	3702	7103	99.42	57.28	1.68

[github-actions]

Test Results

326 tests   320 ✔️  25m 59s ⏱️
110 suites      6 💤
    6 files        0 ❌

Results for commit 7389789.

♻️ This comment has been updated with latest results.

[ch1bo]

Mostly minor things and I'm happy to try this out!

[ch1bo]

I think it's fair to call this now ci-cabal.yaml and the other ci-nix.yaml and have it running along.

Caching seems to look good now as well? Great job!