Cleanup Espresso Config #932

lukeiannucci · 2026-01-20T23:57:54Z

Closes #<ISSUE_NUMBER>

This PR:

We are working towards making the config part of PCR0 hash. In order to do so we are revisiting the config, and what needs to actually be configurable vs what shouldn't be.

Changes to config:
RegisterSignerOpts - This is now all constants as we never really need to change these.
EspressoTxnLimit - This is now a constant.
DangerousConfigHotshotBlockNum - This is removed as we are able to parse hotshot blocks much faster if there is a big gap.
espressoTxnsPollingInterval and espressoTxnsResubmissionInterval - These have been collapsed into one config called espressoTxnsMonitoringInterval. I noticed these were normally the same input in the config.

This PR does not:

Key places to review:

gemini-code-assist · 2026-01-20T23:58:13Z

Summary of Changes

Hello @lukeiannucci, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly refactors the configuration management for Espresso TEE service registration. It centralizes various transaction and retry-related parameters into global constants, thereby eliminating the need to pass these configurations through multiple structs and function signatures. This change simplifies the API of the TEE key manager and verifier components, reduces configuration overhead in related modules like the batch poster and Caff node, and promotes a more consistent approach to these operational settings.

Highlights

Centralized TEE Service Configuration: Configuration parameters for Espresso TEE service registration, such as transaction wait time, retry logic, and gas limit buffer, have been moved from various configurable structs to global constants within espressotee/espresso_tee_helpers.go.
Simplified API for TEE Key Manager and Verifier: The EspressoKeyManager and EspressoTEEVerifier interfaces and their implementations have been streamlined by removing the EspressoRegisterServiceConfig and EspressoRegisterServiceOpts parameters from their constructors and method calls.
Reduced Configuration Overhead: The RegisterServiceConfig field has been removed from EspressoBatchPosterConfig and EspressoCaffNodeConfig, simplifying the configuration structures for these components and reducing the need to pass these settings explicitly.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request successfully cleans up the Espresso configuration by removing the EspressoRegisterServiceConfig struct and its associated fields and options. The parameters previously managed by this struct are now defined as constants within espressotee/espresso_tee_helpers.go. This change streamlines the configuration process and removes unused code, contributing to a cleaner and more maintainable codebase. All modifications are consistent across the affected files, including arbnode/batch_poster.go, arbnode/espresso_batch_poster_config.go, arbnode/espresso_caff_node.go, espresso/key-manager/key_manager.go, espressotee/espresso_tee_helpers.go, espressotee/espresso_verifier.go, and system_tests/caff_node_test.go. The removal of unused imports also contributes to the overall cleanup. The changes are well-executed and align with the pull request's objective.

github-actions · 2026-01-21T00:41:08Z

✅ All tests successful. No failed tests were found.

📣 Thoughts on this report? Let Codecov know! | Powered by Codecov

lukeiannucci

/gemini review

varun-doshi · 2026-01-22T05:58:27Z

arbnode/batch_poster.go

-					panic(err)
+				if registrationFailCount > espressotee.EspressoMaxRetries {
+					log.Crit("Espresso signer registration failed 5 times consecutively. Stopping.", "err", err)
+					b.fatalErrChan <- err


we would need to handle this error in node.go specifically on this line
So instead it should be:

if n.BatchPoster != nil { err = n.BatchPoster.Start(ctx) if err != nil { return fmt.Errorf("failed to start batch poster: %w", err) } }

I am not convinced this is needed, when we start the batch poster it is launching a new thread which is trying to post a batch and register the signer address, and it may take some time for it to register the address (say for example 5 retries, 20 seconds each). So, while this is going on we want to bring up any other components and we are never waiting around to see if batchPoster.Start(ctx) errors out. When we trigger the error through the channel, it should simply stop all components that are running. Does this make sense? What do others think?

hmm yes that makes sense to me. I just wasn't sure if fataErrChan getting populated would automatically trigger a shutdown

its a little bit confusing so here it defers the StopAndWait()

nitro-espresso-integration/cmd/nitro/nitro.go

Line 772 in e8d0468

deferFuncs = []func(){func() { currentNode.StopAndWait() }}

so when this method exits this is called it stops all the components that were started:

nitro-espresso-integration/arbnode/node.go

Line 1562 in e8d0468

func (n *Node) StopAndWait() {

it only gets exited through a sig int or this channel:

nitro-espresso-integration/cmd/nitro/nitro.go

Line 794 in e8d0468

case err = <-fatalErrChan:

I also browsed through the codebase and sometimes as you mention some components return an error and sometimes they dont when using the fatalChanErr. So i think it is fine to not return an error here, but we can see what the team thinks!

There is a reason why in Start() we dont wait for it to register the signer and do it in a background thread. It is a bit long to explain but essentially there can be dependencies on other components being started, in order for the registration to complete successfully. Does this make sense?

Ahh understood
Thank you for clearing this up

arbnode/batch_poster.go

varun-doshi

It would also be nice to make the appropriate changes to the migrate-config-file.py according to the config changes you made but I would suggest waiting till this PR gets merged (which should be done by today I believe)

varun-doshi

~~Also I may be wrong but we need to update the BatchPoster.Start to return an error and use CallIterativelySafe instead of CallIteratively~~

~~Essentially, the structure needs to be similar to what ForceInclusionChecker does~~

Not required

Sneh1999 · 2026-01-22T20:27:20Z

arbnode/espresso_transaction_streamer_helpers.go

-		TxnsResubmissionInterval:              DefaultEspressoBatchPosterConfig.TxnsResubmissionInterval,
-		MaxTransactionSize:                    DefaultEspressoBatchPosterConfig.TxSizeLimit,
-		ResubmitEspressoTxDeadline:            DefaultEspressoBatchPosterConfig.ResubmitEspressoTxDeadline,
+		TxnsResubmissionInterval:              DefaultEspressoBatchPosterConfig.TxnsMonitoringInterval,


why cant we also combine these two?
f.Duration(prefix+".txns-sending-interval", DefaultEspressoBatchPosterConfig.TxnsSendingInterval, "interval between sending transactions to Espresso Network")
f.Duration(prefix+".txns-monitoring-interval", DefaultEspressoBatchPosterConfig.TxnsMonitoringInterval, "time threshold after which a transaction will be automatically resubmitted if no response is received")

We probably can, I just noticed in the configs we currently have they tended to be different. I can look more into this. I think its a good idea to try and collapse them even more

i ended up making combining
TxnsSendingInterval and TxnsPollingInterval as these were always similar. The TxnsResubmissionInterval, i left different because if i understand it resubmits a transaction only if its not finalized by hotshot yet. For fast chains we would make TxnsSendingInterval and TxnsPollingInterval fast, like 125ms, so i dont think it makes sense to merge TxnsResubmissionInterval to the same value as hotshot may take a couple seconds - if not more for finalization

Sneh1999 · 2026-01-22T20:27:56Z

espresso/key-manager/key_manager.go

 	serviceType             espressotee.ServiceType
-	registerSignerOpts      espressotee.EspressoRegisterServiceOpts
 	userDataAttestationFile string
 	quoteFile               string


quoteFile and useDataAttestationFile can also be hardcoded I think because they are only needed for SGX

6813fb9

I was able to do it in here, however this cause some problems as the tests were hard coded for tee type "SGX" and since the default for these fields were userDataAttestationFile and quoteFile were empty strings it would work as getAttestationQuote was returning empty bytes and not erroring out. I tried to change the tests to tee type "TESTS" but then the mock contract would revert the calls because enum was 2, and it is only aware of 0 and 1 (NITRO, SGX). Anyway I was able to hack it together, I think a more proper fix would be for mock contract to support TESTS as well.

instead of hacking can you just create a PR for mock contracts lol

yea im getting to it lmao

this should fix it:
dc42a15

ugh, it actually is harder than i thought as the nitro-contracts do this:

(uint256 hotshotHeight, bytes memory signature, IEspressoTEEVerifier.TeeType teeType) = abi.decode( espressoMetadata, (uint256, bytes, IEspressoTEEVerifier.TeeType) );

meaning they only support espressotee enums, not mockespressotee enums

I see makes sense!

Sneh1999 · 2026-01-22T20:30:34Z

arbnode/espresso_caff_node.go

+	SGXVerifierAddr        string                  `koanf:"sgx-verifier-addr"`
+	BatchPosterAddr        string                  `koanf:"batch-poster-addr"`
+	RecordPerformance      bool                    `koanf:"record-performance"`
+	WaitForFinalization    bool                    `koanf:"wait-for-finalization"`


can we like combine WaitForFinalization WaitForConfirmations and RequiredBlockDepth I feel like this can be like an enum or something? like if we are suppling RequiredBlockDepth its basically waiting for confirmations

Yea, I didnt look to much into caff node changes as that was a bit outside of my area of expertise, but i will do so now.

1aba34a

checkout what i did here i removed WaitForConfirmations and if RequiredBlockDepth is greater than 0 we go inside the WaitForConfirmations code blocks

Sneh1999 · 2026-01-22T20:31:11Z

arbnode/espresso_caff_node.go

+	WaitForFinalization    bool                    `koanf:"wait-for-finalization"`
+	WaitForConfirmations   bool                    `koanf:"wait-for-confirmations"`
+	RequiredBlockDepth     uint64                  `koanf:"required-block-depth"`
+	BlocksToRead           uint64                  `koanf:"blocks-to-read"`


can we make this constant?

Which field? I can look into it

BlocksToRead

espresso/key-manager/key_manager.go

varun-doshi · 2026-01-27T14:42:54Z

The script optimisations PR was just merged. You can make appropriate changes to the script now 🫡

lukeiannucci · 2026-01-28T16:46:47Z

The script optimisations PR was just merged. You can make appropriate changes to the script now 🫡

Thanks @varun-doshi let me know what you think of this commit:
ffb6592

varun-doshi

LGTM

github-actions · 2026-01-28T21:16:55Z

Backport failed for integration, because it was unable to cherry-pick the commit(s).

Please cherry-pick the changes locally and resolve any conflicts.

git fetch origin integration
git worktree add -d .worktree/backport-932-to-integration origin/integration
cd .worktree/backport-932-to-integration
git switch --create backport-932-to-integration
git cherry-pick -x 9837028f180efa8e763b6ab3500216cdca3e5ccb

github-actions · 2026-01-28T21:16:57Z

Backport failed for integration-v3.5.6, because it was unable to cherry-pick the commit(s).

Please cherry-pick the changes locally and resolve any conflicts.

git fetch origin integration-v3.5.6
git worktree add -d .worktree/backport-932-to-integration-v3.5.6 origin/integration-v3.5.6
cd .worktree/backport-932-to-integration-v3.5.6
git switch --create backport-932-to-integration-v3.5.6
git cherry-pick -x 9837028f180efa8e763b6ab3500216cdca3e5ccb

github-actions · 2026-01-28T21:16:59Z

Backport failed for celestia-v3.5.6, because it was unable to cherry-pick the commit(s).

Please cherry-pick the changes locally and resolve any conflicts.

git fetch origin celestia-v3.5.6
git worktree add -d .worktree/backport-932-to-celestia-v3.5.6 origin/celestia-v3.5.6
cd .worktree/backport-932-to-celestia-v3.5.6
git switch --create backport-932-to-celestia-v3.5.6
git cherry-pick -x 9837028f180efa8e763b6ab3500216cdca3e5ccb

github-actions · 2026-01-28T21:17:00Z

Backport failed for celestia-v3.9.2, because it was unable to cherry-pick the commit(s).

Please cherry-pick the changes locally and resolve any conflicts.

git fetch origin celestia-v3.9.2
git worktree add -d .worktree/backport-932-to-celestia-v3.9.2 origin/celestia-v3.9.2
cd .worktree/backport-932-to-celestia-v3.9.2
git switch --create backport-932-to-celestia-v3.9.2
git cherry-pick -x 9837028f180efa8e763b6ab3500216cdca3e5ccb

* start * more cleanup * remove unused * more fixes * add fatal err chan to tests * start addressing comments * address more comments * add back tx deadline * fix mock * hack * re add * fix script * fixes

* Cleanup Espresso Config (#932) * start * more cleanup * remove unused * more fixes * add fatal err chan to tests * start addressing comments * address more comments * add back tx deadline * fix mock * hack * re add * fix script * fixes * fixes * submodule

start

d5f4983

lukeiannucci changed the title ~~start~~ Cleanup Espresso Config Jan 20, 2026

gemini-code-assist bot reviewed Jan 21, 2026

View reviewed changes

lukeiannucci added 4 commits January 21, 2026 13:32

Merge branch 'integration-v3.9.2' into li/config-cleanup

2cf230f

more cleanup

01fe91e

remove unused

961696c

more fixes

57e1b2f

lukeiannucci marked this pull request as ready for review January 21, 2026 20:40

lukeiannucci requested review from ImJeremyHe, Sneh1999, jbearer, jjeangal, philippecamacho, varun-doshi and zacshowa as code owners January 21, 2026 20:40

lukeiannucci commented Jan 21, 2026

View reviewed changes

varun-doshi reviewed Jan 22, 2026

View reviewed changes

add fatal err chan to tests

29d4208

Sneh1999 reviewed Jan 22, 2026

View reviewed changes

lukeiannucci added 6 commits January 23, 2026 11:50

start addressing comments

6813fb9

Merge branch 'integration-v3.9.2' into li/config-cleanup

10472d7

address more comments

1aba34a

add back tx deadline

34d51cc

fix mock

dc42a15

hack

c914b3e

re add

0a4c072

Sneh1999 approved these changes Jan 28, 2026

View reviewed changes

lukeiannucci added 3 commits January 28, 2026 11:24

fix script

b644160

Merge branch 'integration-v3.9.2' into li/config-cleanup

2c5cf02

fixes

ffb6592

lukeiannucci requested a review from varun-doshi January 28, 2026 16:46

varun-doshi approved these changes Jan 28, 2026

View reviewed changes

lukeiannucci added backport integration backport integration-v3.5.6 backport celestia-v3.5.6 backport celestia-v3.9.2 labels Jan 28, 2026

lukeiannucci merged commit 9837028 into integration-v3.9.2 Jan 28, 2026
33 of 34 checks passed

lukeiannucci deleted the li/config-cleanup branch January 28, 2026 21:16

lukeiannucci mentioned this pull request Jan 28, 2026

Backport #953

Merged

Cleanup Espresso Config #932

Cleanup Espresso Config #932

Uh oh!

Conversation

lukeiannucci commented Jan 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

This PR:

This PR does not:

Key places to review:

Uh oh!

gemini-code-assist bot commented Jan 20, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

github-actions bot commented Jan 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lukeiannucci left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

varun-doshi Jan 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lukeiannucci Jan 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

varun-doshi left a comment

Choose a reason for hiding this comment

Uh oh!

varun-doshi left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lukeiannucci Jan 28, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lukeiannucci Jan 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lukeiannucci commented Jan 20, 2026 •

edited

Loading

github-actions bot commented Jan 21, 2026 •

edited

Loading

lukeiannucci left a comment •

edited

Loading

varun-doshi Jan 22, 2026 •

edited

Loading

lukeiannucci Jan 22, 2026 •

edited

Loading

varun-doshi left a comment •

edited

Loading

lukeiannucci Jan 28, 2026 •

edited

Loading

lukeiannucci Jan 23, 2026 •

edited

Loading