Bump github.com/quic-go/quic-go from 0.46.0 to 0.57.0 #300
Conversation
This file contains hidden or 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
…optimism#405) * github: Detect OP_GETH_BASE_BRANCH in update-geth workflow Minor additional changes: * use a more informative job id than "job_id" * add quotes to silence shellcheck warnings Closes celo-org/celo-blockchain-planning#933 * github: Fix ambiguous ls-remote match The previous command yielded multiple results: ``` > git ls-remote https://github.com/celo-org/op-geth/ "celo-rebase-14" 18fdf55a858a46e5838c72f0e1eddcae00ced616 refs/heads/Kourin1996/celo-rebase-14 df2ea968968d4910ce21b44280a0077a54f73d3f refs/heads/celo-rebase-14 ``` That led to a malformed GITHUB_OUTPUT and caused job errors like ``` Error: Unable to process file command 'output' successfully. Error: Invalid format 'df2ea968968d4910ce21b44280a0077a54f73d3f' ```
dependabot
bot
added
dependencies
Initial implementation of Espresso integration by using Espresso as an AltDA layer.
Implement getting AWS Nitro attestations and commitmen signing in the batcher
In order to be able to verify that a transaction coming from Espresso's finalized transaction list was originally sourced from the op-batcher, we need a way to verify that the transaction did actually come from the op-batcher. To facilitate this we want the transaction submitted to espresso to be signed by the op-batcher itself. The majority of this change is a result of attempting to forward the signing functions to the place where we need them, while also maintaining no ability to have direct memory access to any potentially stored private key. This change signs the payload with the desired private key before it is submitted to Espresso. Update comments on `clientSigner` and `privateKeySigner` The comment on `privateKeySigner` is for different code altogether. It should be replaced with a comment that applies to `privateKeySigner` itself. Additionally the comments could use some more explicit purpose. This change updates the comments on `privateKeySigner` and `clientSigner` that help to explain what they are intended for. Fix ChainSigner not being set for Config and DriverSetup The ChainSigner is created, but it's never assigned to Config and for DriverSetup. This leads to null pointer calls when trying to sign the payload for the chain. Clean up comment that had an extra "creates a"
The rebase from celo-tip made the celo-integration branch go back a version on`github.com/minio/minio-go/v7` from `v7.0.85` to `v7.0.84`. Additionally, the devnet configuration files ended up dropping the `participants` configurations. This commit updates the version to what it was, and re-adds the participants configurations in the devnet files.
Pre-download rust dependency of Espresso Go SDK in CI builds
Removes reliance on AltDA for Espresso integration * Allows for simultaneous usage of Espresso and an AltDA provider * Allows for immediate posting of blocks to Espresso instead of waiting for new frames on a channel * When not using an alternative DA layer, L1 now receives full batch data instead of just the commitments
…`espresso-dev-node`
* Same logic for the Espresso Streamer used in the batcher and derivation pipeline * Separate the interaction between Espresso nodes and the handling of batches * Several TODOs / improvements will be addressed in another PR.
Reduces code duplication and move more code to a separate package by making Batch a type parameter of the streamer
* Document how to make changes to the kona repository and propagate them. * Reference new docker images for the op-succinct proposer and challenger.
* Remove the superfluous check about the batcher address as now the Batch Inbox contract verifies the sender is legitimate.
* Add support for ZK attestation service * check attestation service url is not nil * upgrade espresso tee verifier contracts * fix contracts * fix merge * fix tests * bring back deploy aws nitro * add support for mock contract * add support for attestation verifier service * fix tee tests * use higher version of github runner * fix tee args * fix tee args * add healthcheck to attestation verifier zk * increase timeout * Invalid attestation test passing * small fixes * fix TestE2eDevnetWithUnattestedBatcherKey * fix health check * fix devnet test * use 127.0.0.1 * fix regex * debug * fix proof generation * debug * fix url * fix url * remove debug logs * resolve based on comments * address comments * update github runner enclave * fix based on suggestions * cleanup logs
* Integrate EigenDA via Docker proxy for AltDA Espresso E2E tests * Scope EigenDA lifecycle to the test and ensure clean startup/teardown * Extract EigenDA Docker port and image into constants
* Make attestation service url optional * fix dasel * fix dasel * update dockerfiles * make attestation service required again
* Withdraw test passing again on devnet. * Faster CI * Deposit into L1 before requesting the withdrawal on L2.
--------- Co-authored-by: Philippe Camacho <[email protected]>
* update logging for the batch * clean up
* Add code sync procedure * Update links * Fix format * Rename files
* update batchAuthenticator according to audit report * gen bindings and fix fast-tests
* Simplify the test as we cannot in practice reduce the window size. --------- Co-authored-by: Philippe Camacho <[email protected]>
* remove warning on every failed tx * reorder the checks
* Fallback mechanism test * Update op-e2e/system/e2esys/setup.go Co-authored-by: Phil <[email protected]> --------- Co-authored-by: Philippe Camacho <[email protected]>
* Address flakiness. * Simplify the code * Fix CI --------- Co-authored-by: Keyao Shen <[email protected]>
* update single run-enclave.sh * remove BATCHER_PRIVATE_KEY * update run-enclave.sh
* Fallback recovery * Add caff node * Suggestions
* Make Attestation Verifier Service optional When the Attestation Verifier Service was added to the integration it fundamentally modified the testing experience, requiring external environment variables to be populated in order to run the tests. Additionally, these environment variable requirements were not documented in the README_ESPRESSO.md file for reference. This change modifies the Attestation Verifier Service setup for the E2E testing environment to make it opt-in instead of being forced to be enabled. Additionally, the Verifier URL is no longer required to run the Batcher. This is a double-edged sword, however, as it means that we could potentially deploy the service without the configuration, and we would potentially be lacking the registered attestation. This may be resolvable with a slight modification to the service configuration, that we would ultimately disable for the E2E testing environment. * Fix misspelling Fix linting error that has caught a misspelling of the work 'Network'. * Modify configuration address to be required from CLI With the change of making the Espresso Attestation Service optional we removed the CLI configuration check that occurs on launch, so that the E2E tests can still be run. This has an unfortunate side-effect of allowing the Batcher to be launched in a state where it is unable to operate as intended due to user error. The only indication being a `WARN` log entry to inform him/her of his/her mistake. This sort of approach is generally discouraged, yet we still need to be able to bypass this check for testing purposes. As a result the `EspressoAttestationService` value has been modified from being a simple `string` to being an interface whose value is inspectable and not allowed to be empty by default. This allows for the test configurations to overwrite this behavior, and allow an optional value in the cases where it is needed. This should preserve the prior behavior of erroring on launch when the parameter is not configured or specified, and should also preserve the new behavior where it is explicitly disabled in tests. * Fix some nil references The EspressoAttestationService configuration value being an interface makes it a `nil`lable value by default. Care needs to be taken when accessing this value an referencing it. This change adds some additional care in referencing the value stored within. * Fix nil access error The `l1Client` being created assumes that the `sys` returned from the call is non-nil before checking the error. This is not guaranteed, and is most likley not ever the case. As a result there is a potential for an error do to attempting an access on a `nil` value. By moving the `l1Client` declaration after the error check, we avoid the potential for this issue. * Apply linting and formatting changes * Fix e2e tests - populate default EspressoAttestationService With the modification of the EspressoAttestationService to an interface instead of an individual value, we need to ensure that the default way of launching the Espresso E2E DevNet results in the value being populated with an empty allowed value. This still allows for extension and override, without requiring the value to be specified, which is our intention. This was missed when adding the capability originally. * Cleanup code practices We have duplicated code that makes the maintenance burden more difficult than it needs to be. In many of these scenarios the code that is duplicated differs by only a single line. Instead of making the system more flexible, we ended up duplicating code paths. This increases the maintenance burden by needing to ensure that these code paths match in every case where they do not differ, yet they are independent of each other. This is not a great approach. Additionally, we end up with multiple starting points for something that should not need them. We also end up storing a configuration that is unnecessary to store. This incurs conditional checks where some are not needed, and ends up making the approach be more confusing than it needs to be. This change aims to replace these approaches with one that adheres to the functional option approach and preserves the existing behavior. * Revert EspressoAttestionService to a `string` As it so happens we rely on the `CLIConfig` for `Espresso`, and the `Batcher` to be serializable. By utilizing an `interface`, we run into trouble doing this. Due this constraint, the `interface` constraint is not feasible. This change reverts the value back to a `string`, which should result in a smaller overall change. It also opts for a private configuration value that is inspectable by the `Verify` check, but not directly configurable. We expose a method to allow for it to be configured, so it can only occur within code within the code base itself. We should only invoke this via Testing where we need the value to be optional. This achieves the same result but in a different way. | NOTE: There may be a better approach to this as well, isntead of having this be a separate field, we could do something akin to sql.NullString, where we encode this value as a Marshable `struct`. The acess pattern becomes different, but we could directly encode the empty allowance into the struct itself. * Add Espresso Attestation Verifier Service to Enclave Test The Enclave tests are currently failing in CI. It is dying due to an error stemming from the lack of the EspressoAttestationService being configured. It is likely that this is required for the Enclave tests specifically. As a result, we need to add and enable it for the enclave tests. * Modify LaunchBatcherInEnclave option The LaunchBatcherInEnclave essentially launches the batcher externally within an enclave. This option actually relies on the Espresso Attestation Verifier Service to be running. This is due to the Espresso Attestation Service only being optional inside of a test environment. When launched externally, the Batcher is no longer considered to be in a "test environment", or configurable for testing. As a result, its configuration **MUST** be something that can actually be resolvable from a CLI launch. Since the Espresso Attestation Verifier Service check is only disabled within the testing environment, this means it **MUST** be enabled in the enclave. For convenience, this option has been added automatically as a part of the LaunchBatcherInEnclave option, since it depends on it. This will minimize accidentl misconfigurations.
* add eigenda_proxy_url to op-batcher-tee * fix the url to post to eigenDA * not hardcoding EIGENDA_PROXY_PORT * fix the block height config
* Add test to check end of channel fallback Asana task: https://app.asana.com/1/1208976916964769/project/1209976130071762/task/1211892212379885?focus=true We need a test to check the fallback Batcher behavior in the event that the Espresso Batcher is able to submit a partial Channel that is im progress. The specific scenario we want to test for is one concerning a multi-frame channel that has had at least part of the full channel submitted to the L1 by the Espresso Batcher, then no more. After which we swap to the Fallback Batcher, and we should be able to pick up the missed / incomplete channel, and complete the transactions. * Rename helper function to match naming pattern * Fix lint issue with not checking error result of wait.For * Commit work in progress multi frame channel efforts * Adjust settings to successfully trigger multi-frame channels After a mob programming session @quentinl was able to help identify a a specific combiniation of parameters to successfully and consistently trigger multi-frames within the Batcher. This condition is a necessary precusor to the test being attempted. This commit updates the test with the information necessary to trigger this condition and sets the necessary test criteria that we are aiming to achieve. * Perform some code cleanup This change does a few things: - Address linting issue causing CI failure - Adjusts some golang forloop usage to be more modern - Adjust function call signatures to remove unused variables * Fix bug tracking unsuccessful frames in test In the `TxManagerIntercept` there is a bug that appends the successful frames to the unsuccessful ones. While this bug isn't great in the information that it taints, it doesn't actually have the large of an impact on the test as a whole, as the resulting failure condition would be triggered regardless. This bug does affect the accurate tracking of failed frames which could be valuable information for inspection. * Update espresso/environment/e2e_helpers.go Co-authored-by: Phil <[email protected]> * Replace Disable Batcher setting references There are a number of places in our testing setup where we are explicitly preventing the Batcher from starting on launch. Instead of rewriting this same option every time we want to use it, we should reference a built in option that we can reference continually. This allows for non-repeated code and improved documentation as to the point and purpose of this option. * Refactor custom wait in test There's a condition being waited on in the switch to fallback batcher test. This wait is useful, and can be reused between tests. But the wait itself is somewhat hiding it's intention by being inline defined within the test itself. We should pull this wait out so it can be easily used, and its intention / purpose can be more easily documented. * Cleanup code reuse in frame decoding When decoding frame information for one of the Batcher fallback tests, there are similar code paths taken that result in most of the code being reused. We should clean up this code reuse so that we don't repeat ourselves in order to avoid diverging logic. Additionally, it allows us to reduce the amount of code needing to be maintained, and more clearly document the intention of the code, and the consistency with how we perform this frame decoding process. * Relocate deferred stop calls The Stop calls should occur as close to the launch of the environment as possible. As a result, any deferred calls to Stop for the system or the Espresso Dev Node should occur as close to their occurence as possible. * Modify Initial L2Verif wait to be longer With the specific Frame and Channel settings being specified in the `TestFallbackMechanismIntegrationTestChannelNotClosed` test, the initial startup check for the L2 Verifier is failing. This is due to our settings requiring the Verifier process to take a bit longer than normal. In general, we want to give it more time, but the time frame for the failure is hard-coded in the `wait` function being utilized. While we **could** add a simple `time.Sleep`, and this would work, this is generally a bad appraoch as it just adds an unchecked delay. Instead, we opt to utilize a simple `retry` for up to `n` times. In this case, we only need to wait up to `3x` the normal time, so ensure that we perform at least `3` times. * Fix failure in Batcher Fallback test The TestFallbackMechanismIntegrationTestChannelNotClosed test fails locally without stopping, in spite of the overall time limit being specified on the test. After some troubleshooting and debugging, We were able to chase down the cause to be due to the `RunSimpleMultiTransactions`. It's unclear as to why this was causing the process to hang for as long as it was. It seemed to not be handling timeout errors well for some reason. Either way, we fority this helper by setting an explicit time limit on it, and referncing the context whenever we're performing channel operations. This should allow the channel operations themselves not to block and hang the test. After this modification we were able to determine that this process was failing due to insufficient gas being provided. For some reason when running the transactions through this mechanism, we require even more gas than we're normally need. This seems a bit odd, perhaps it has to do with the differences in the transaction construction. In any case, we up the gas being provided so that this becomes a non-issue. * Fix linting issues * Update espresso/environment/tx_helpers.go Co-authored-by: Phil <[email protected]> * Correct failure vs success in Send The triggered conditions for failures and successes are backwards in the `Send` method of `TxMangerIntercept`. Their specific frame markers should be switched. * Update espresso/environment/14_batcher_fallback_test.go Co-authored-by: Phil <[email protected]> --------- Co-authored-by: Phil <[email protected]>
* Check if the batcher is active before publishing to L1/DA * fix readme lint * more lint fixes * check batcher contract * Fix endless warning * add batch authenticator address to rollup config * handle contract undeployed error * attempt test in CI * add test to CI * Revert "add test to CI" This reverts commit 2a9678a. * add test to CI * remove jg/ from branches * attempt to clean up and make the test more reliable * fix ci error WaitUntilSafe undefined * revert 07a82bf * Fix `anvil_setBalance` not found error * Simplify isActive check * add batcher-active-publish-only to devnet tests justfile * - simplify test, one less batcher switch - increase timeouts for devnet test * Cleaned up the code, raise tx waiting time to 60s * Brought back original timeouts * started fallback batcher up + lint fix docker compose file * Ensure that in Espresso mode the batch authenticator address is set. * Removing all changes to driver.go and the tests are still passing. --------- Co-authored-by: Philippe Camacho <[email protected]>
* Remove pre authenticated batcher * fix test
Bumps [github.com/quic-go/quic-go](https://github.com/quic-go/quic-go) from 0.46.0 to 0.57.0. - [Release notes](https://github.com/quic-go/quic-go/releases) - [Commits](quic-go/quic-go@v0.46.0...v0.57.0) --- updated-dependencies: - dependency-name: github.com/quic-go/quic-go dependency-version: 0.57.0 dependency-type: indirect ... Signed-off-by: dependabot[bot] <[email protected]>
dependabot
bot
changed the base branch from
celo-integration-rebase-14.1
to
celo-integration-rebase-14.2
dependabot
bot
force-pushed
the
dependabot/go_modules/github.com/quic-go/quic-go-0.57.0
branch
from
c55817b to
3e58a02
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
13 participants
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.
Bumps github.com/quic-go/quic-go from 0.46.0 to 0.57.0.
Release notes
Sourced from github.com/quic-go/quic-go's releases.
... (truncated)
Commits
5b2d212http3: limit size of decompressed headers (#5452)e80b378qlogwriter: use synctest to make tests deterministic (#5454)d43c589README: add nodepass to list of projects (#5448)ca2835ddon’t arm connection timer for connection ID retirement (#5449)e84ebaeackhandler: don’t generate an immediate ACK for the first packet (#5447)d4d168fadd documentation for Conn.NextConnection (#5442)4cdebbehttp3: use QPACK_DECOMPRESSION_FAILED for QPACK errors (#5439)b7886d5update qpack to v0.6.0 (#5434)2fc9705http3: add a benchmark for header parsing (#5435)dafdd6fhttp3: make Transport.MaxResponseBytes an int (#5433)You can trigger a rebase of this PR by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot mergewill merge this PR after your CI passes on it@dependabot squash and mergewill squash and merge this PR after your CI passes on it@dependabot cancel mergewill cancel a previously requested merge and block automerging@dependabot reopenwill reopen this PR if it is closed@dependabot closewill close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)You can disable automated security fix PRs for this repo from the Security Alerts page.