fix(prover): Remove FriProtocolVersionId #1510
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This is PR 2 of many to move us to safe Protocol Upgrades (phase 1).
`FriProtocolVersionId` was introduced as a cludge to serve as
communication between core (which was saved from L1 straight to DB) and
prover (which was manually configured from env vars). Whilst in theory
it's possible to have an upgrade on core side without affecting prover,
I don't envision this happening any time soon. Right now we maintain
2 different protocol numbers that cause confusion. This PR unifies the
protocol ids.
Other relevant notes:
- this is not exactly backwards compatible -- we'll need to deploy both
server and prover at the same time, otherwise provers will go in crash
loop. We can make it backwards compatible, but at our track record of
deployments (and speed of deployments), I'm concerned we will slow
everything down for no real reason. I see very small risks here
- database tables in prover's end should be renamed to protocol_version
as well (to be covered in next PR)
- all our code base assumes the old prover's interface to hashes which
is confusing and unnecessary. I propose we address those with
following PR. The points there are (to be done on both databases and
configs):
- rename `recursion_scheduler_level_vk_hash` to
`snark_wrapper_vk_hash`. This is what it represents today. I
believe we can't change `recursion_scheduler_level_vk_hash` without
changing `snark_wrapper_vk_hash`.
- delete 'recursion_circuits_set_vks_hash`; it's all 0.
- (prover only) rename table to `protocol_versions` from
`prover_fri_protocol_versions`
EmilLuta
requested review from
mm-zk,
Artemka374,
Deniallugo,
popzxc,
ColoCarletti and
ilitteri
and removed request for
mm-zk
mm-zk
approved these changes
Mar 27, 2024
|
where does backwards incompatibility come from? (as AFAIK we actually used the same value for both) |
Deniallugo
approved these changes
Mar 27, 2024
|
ok |
This was referenced Mar 27, 2024
RomanBrodetski
pushed a commit
that referenced
this pull request
Mar 28, 2024
🤖 I have created a release *beep* *boop* --- ## [22.1.0](core-v22.0.0...core-v22.1.0) (2024-03-28) ### Features * Drop prover tables in core database ([#1436](#1436)) ([0d78122](0d78122)) * **en:** consistency checker persistent cursor ([#1466](#1466)) ([03496e6](03496e6)) * **en:** Make snapshot syncing future-proof ([#1441](#1441)) ([8c26a7a](8c26a7a)) * **genesis:** Using genesis config only during the genesis ([#1423](#1423)) ([4b634fd](4b634fd)) * **node_framework:** Add a task to handle sigint ([#1471](#1471)) ([2ba6527](2ba6527)) * **node-framework:** Add circuit breaker checker layer to framework ([#1452](#1452)) ([2c7a6bf](2c7a6bf)) * **prover:** export prover traces through OTLP ([#1427](#1427)) ([16dce75](16dce75)) * sigint initialization only after snapshots is applied ([#1356](#1356)) ([c7c7356](c7c7356)) * Split witness generator timeout configs by round ([#1505](#1505)) ([8074d01](8074d01)) * **state-keeper:** implement asynchronous RocksDB cache ([#1256](#1256)) ([da41f63](da41f63)) * **state-keeper:** Refactor persistence in `StateKeeper` ([#1411](#1411)) ([e26091a](e26091a)) * **state-keeper:** Remove `WitnessBlockState` generation from state keeper ([#1507](#1507)) ([8ae0355](8ae0355)) * Switch contract verification API to axum and get rid of actix-web usage ([#1467](#1467)) ([e7a9d61](e7a9d61)) ### Bug Fixes * **api:** `filters_disabled` should only affect HTTP endpoints ([#1493](#1493)) ([8720568](8720568)) * **api:** Fix API server shutdown flow ([#1425](#1425)) ([780f6b0](780f6b0)) * **prover:** Remove FriProtocolVersionId ([#1510](#1510)) ([6aa51b0](6aa51b0)) * **prover:** Remove redundant LoadingMode ([#1496](#1496)) ([e7583f4](e7583f4)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
github-merge-queue bot
pushed a commit
that referenced
this pull request
Mar 28, 2024
🤖 I have created a release *beep* *boop* --- ## [12.2.0](prover-v12.1.0...prover-v12.2.0) (2024-03-28) ### Features * **api:** introduce mempool cache ([#1460](#1460)) ([c5d6c4b](c5d6c4b)) * **commitment-generator:** `events_queue` shadow mode ([#1138](#1138)) ([9bb47fa](9bb47fa)) * Drop prover tables in core database ([#1436](#1436)) ([0d78122](0d78122)) * Follow-up for DAL split ([#1464](#1464)) ([c072288](c072288)) * **prover:** export prover traces through OTLP ([#1427](#1427)) ([16dce75](16dce75)) * **prover:** File-info tool to help prover debugging ([#1216](#1216)) ([9759907](9759907)) * Separate Prover and Server DAL ([#1334](#1334)) ([103a56b](103a56b)) * support running consensus from snapshot (BFT-418) ([#1429](#1429)) ([f9f4d38](f9f4d38)) ### Bug Fixes * **api:** Fix API server shutdown flow ([#1425](#1425)) ([780f6b0](780f6b0)) * **prover:** Remove FriProtocolVersionId ([#1510](#1510)) ([6aa51b0](6aa51b0)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). Co-authored-by: Roman Brodetski <rb@matterlabs.dev>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is PR 2 of many to move us to safe Protocol Upgrades (phase 1).
FriProtocolVersionIdwas introduced as a cludge to serve as communication between core (which was saved from L1 straight to DB) and prover (which was manually configured from env vars). Whilst in theory it's possible to have an upgrade on core side without affecting prover, I don't envision this happening any time soon. Right now we maintain 2 different protocol numbers that cause confusion. This PR unifies the protocol ids.Other relevant notes:
recursion_scheduler_level_vk_hashtosnark_wrapper_vk_hash. This is what it represents today. I believe we can't changerecursion_scheduler_level_vk_hashwithout changingsnark_wrapper_vk_hash.recursion_circuits_set_vks_hash; it's all 0.