Repository navigation
ConsensusQueue: wire no-op ReportingPlugin - #23901
Conversation
|
✅ No conflicts with other open PRs targeting |
|
I see you updated files related to
|
78e9b52 to
e04cd2d
Compare
e04cd2d to
1c33ed0
Compare
| type contractTransmitter struct { | ||
| lggr logger.Logger | ||
| fromAccount ocrtypes.Account | ||
| eventSink triggers.EventSink |
There was a problem hiding this comment.
so at the very least I would expect the ability to map a workflow ID to an event sink. we chose to have just the single engine registry and this is wrapped by the coordinator. If we wire as-is, we need a TODO to come back and address this dependency
There was a problem hiding this comment.
Created this ticket for this 👉🏼 https://smartcontract-it.atlassian.net/browse/CRE-7332
|
CORA - Analysis SkippedReason: The number of code owners (4) is less than the minimum required (5) and/or the number of CODEOWNERS entries with changed files (4) is less than the minimum required (2). |
3d49d6a to
ec7c4a4
Compare
ec7c4a4 to
8f1afd8
Compare
…OCR3 job with contractQualifier: consensus-queue
…DefaultKey, this solves the zero configDigest error
… of the still-nil embedded Service
8ee18a7 to
034798a
Compare
034798a to
980d4c7
Compare
| @@ -0,0 +1,14 @@ | |||
| type = "offchainreporting2" | |||
There was a problem hiding this comment.
Is it worth chatting with @bolekk / @vyzaldysanchez to see whether we can just launch this with the cap reg, and so avoid the job spec?
There was a problem hiding this comment.
(Obvs doesn't need to block this PR)
There was a problem hiding this comment.
registry driven launch ticket 👉🏼 https://smartcontract-it.atlassian.net/browse/CRE-7342
| return nil | ||
| } | ||
|
|
||
| // deployContract deploys the OCR3 contract the job spec uses as its contractID. |
There was a problem hiding this comment.
Why does a contract still needed? This is just for local CRE?
There was a problem hiding this comment.
not only local CRE, the template has this contractID that needs to be filled. Its a contract because the way the ProposeJobSpec tooling works is that it resolves the contractID from filtering on contractQualifier
Its also used when building relayer.NewPluginProvider
If we move to the registry-driven launch @cedric-cordenier mentioned in the other comment this contract and jobspec go away.
| } | ||
|
|
||
| // Vault and DKG carry their plugin config in the OCR offchain config. | ||
| // Other OCR 3.1 plugins read it from the job spec, so it stays empty here. |
There was a problem hiding this comment.
I don't think this works -- we want the plugin config to be onchain so it's part of the config digest; this is a safety mechanism that ensures domain separation in case of diverging configs running
There was a problem hiding this comment.
ok, this is because we don't want to try to reach consensus with different configs (possibly different behavior) and to ensure that we use the configDigest that is carried on each consensus round and nodes ignore if they differ from the one they have, right? yeah makes sense to me. For the nop implementation its not needed I think, maybe tackling this in a separate one?
There was a problem hiding this comment.
ticket to read config from config-digest protected ocr config 👉🏼 https://smartcontract-it.atlassian.net/browse/CRE-7336
The plugin produces no reports, so libocr never calls Transmit. Event delivery will happen in the plugin's Committed step instead.




https://smartcontract-it.atlassian.net/browse/CRE-6009
Requires:
Supports: