8 Commits (22871425d6119761f7cc33d8255eb71e31c72a50)

Author	SHA1	Message	Date
Roman V	8fee6aedc5	fix: GCS correctly splitting storage location into bucket and folder (if present) (#4539) ... ### Description Previously Checkpoint Syncer wasn't correctly parsing the storage location string. If no folder was provided, it treated the object name as folder, e.g. reading `gs://bucket-name/gcsLatestIndexKey` initialized StorageClient with bucket `bucket-name` and folder `gcsLatestIndexKey`. This was due to incorrect logic for splitting the string by `/` that expected a folder to be present. In this PR I fixed this logic, treating a location with only one `/` as _bucket/object_ and with two `/`s as _bucket/folder/object_. ### Backward compatibility Yes ### Testing Manual --------- Co-authored-by: Trevor Porter <tkporter4@gmail.com>	2 months ago
Danil Nemirovsky	5cb1787044	feat: E2E Test Cosmos Scraper Invariant (#4536) ... ### Description - Add metrics based Scraper invariant into E2E Test for Cosmos - Report error if transaction contains multiple contract execution messages - Report error if fee are expressed in unsupported denomination ### Related issues - Successfully scrape Cosmos chains in e2e test #3355 ### Backward compatibility Yes ### Testing E2E Testing --------- Co-authored-by: Danil Nemirovsky <4614623+ameten@users.noreply.github.com>	2 months ago
Daniel Savu	4508b4fed1	fix: don't reuse prefix key in relayer (#4560) ... ### Description This PR fixes a bug @tkporter flagged, which would cause agents to reindex every igp payment on restart (but luckily with no impact on message processing). Unfortunately the same DB prefix key was used to store different items: `gas_payment_by_sequence` and `gas_payment_block_by_sequence`, with the latter overwriting the former. Sequence indexing relies on `gas_payment_by_sequence` to know whether a certain payment was already indexed - if not, it queries an RPC for it and stores it in the db. But because `gas_payment_by_sequence` would get overwritten and thus corrupted, checking whether a certain `gas_payment_by_sequence` is already in the db would fail with a decoding error, and because errors are treated as if the payment just wasn't indexed yet, the RPC call would be made again. Fortunately no migration is required, since just re-running the relayer will cause it to overwrite the old corrupted entries with correct ones, and would also create new prefix entries for the igp blocks.	2 months ago
Daniel Savu	b872d635e1	chore: include txids in indexing logs; add vscode config for rust-ana… (#4537) ... - adds support for rust-analyzer with the new workspace split - includes txids in agent indexing logs, to more easily link message_ids to their txid. @Mo-Hussain and I spent quite some time today trying to find the corresponding txid	2 months ago
Danil Nemirovsky	4151317b00	feat: Fetch logs by transaction hash (#4510) ... ### Description Cosmos indexers now can fetch logs by transaction hash. It allows to reprocess transaction which may be missed by range based indexers. ### Drive-by changes Quite a bit of refactoring ### Related issues Contributes into issue #4300 ### Backward compatibility Yes ### Testing E2E Tests --------- Co-authored-by: Danil Nemirovsky <4614623+ameten@users.noreply.github.com>	2 months ago
Trevor Porter	57b0735472	feat: support multiple validator sets for a single app context in the relayer's observed validator metrics (#4495) ... ### Description - Fixes https://github.com/hyperlane-xyz/hyperlane-monorepo/issues/4030 - The general idea of this change is that if the function `set_validator_latest_checkpoints` is called multiple times in a single message's metadata building, the reset of metrics relating to previously set validators now won't always occur. Atm, if two different multisig ISM validator sets are aggregated, they race each other to update the metric. With this change I expect the contention here to not occur anymore, and to still be able to handle changes in validator sets - This is a bit of a quick fix rather than something super robust tbh - I think in a vacuum I'd prefer something not dependent on time but instead dependent on a context of what message is being processed or something. But at the moment we're seeing very noisy alerts because of this issue, so would like to fix it ASAP. ### Drive-by changes <!-- Are there any minor or drive-by changes also included? --> ### Related issues <!-- - Fixes #[issue number here] --> ### Backward compatibility <!-- Are these changes backward compatible? Are there any infrastructure implications, e.g. changes that would prohibit deploying older commits using this infra tooling? Yes/No --> ### Testing <!-- What kind of testing have these changes undergone? None/Manual/Unit Tests -->	2 months ago
Paul Balaji	1ab8751afb	fix: docker agent build (#4481) ... fix: docker agent build - update image from 1.72.1 -> 1.80.1 drive-by: update more `rust/` --> `rust/main/` drive-by: consolidate e2e non-cosmwasm cache, multiple cache steps per job can get in the way of each other --------- Signed-off-by: pbio <10051819+paulbalaji@users.noreply.github.com>	3 months ago
Mantas-M	096aeff13e	feat: Splitting `rust` workspace into `main` and `sealevel` (#4438) ... ### Description The single rust workspace has been split into 2 workspaces. The sealevel workspace includes the contents of the previous `sealevel` folder and the main workspace includes everything else. The workspaces differ in rust versions as the sealevel has the old one to ensure compatability, the main workspace has and updated `rustc` version of `1.80.1`. Some dependencies and sub-dependencies have been locked in versions of the old pre-split workspace to ensure no breaking code changes. The rust CI workflow has been updated to include both workspaces in the tests. ### Drive-by changes N/A ### Related issues Relates to https://github.com/hyperlane-xyz/hyperlane-monorepo/issues/4361 ### Backward compatibility No. ### Testing Rust Unit tests from the CI	3 months ago