Conversation
packages/kusama/src/__snapshots__/assetHubKusama.multisig.e2e.test.ts.snap
Show resolved
Hide resolved
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
There was a problem hiding this comment.
The PR updates block numbers and endpoints, which is generally fine. However, there is a significant inconsistency between the snapshot changes and the visible code changes. Specifically, a snapshot for collectivesPolkadot implies changes to the shared accounts test logic that are not present in the PR.
packages/polkadot/src/__snapshots__/collectivesPolkadot.accounts.e2e.test.ts.snap
Outdated
Show resolved
Hide resolved
Some were wrong due to copy-pasting.
rockbmb
added
e2e tests
Previously, it was assuemd they would be filtered post-AHM, but this is not the case.
|
RPCs are stalling frequently, especially for Polkadot Hub; changing endpoints didn't help. Will try different combinations of endpoints. |
rockbmb
left a comment
rockbmb
left a comment
There was a problem hiding this comment.
I believe the failures are because the large number of tests running concurrently overwhelm the selected endpoints; especially if many tests for the same particular chain are run at once.
I will look into this, as polkadot-fellows/runtimes#1010 will need it too.
The tentative plan is to query all chains once at KNOWN_GOOD_BLOCK_NUMBERS to setup a DB at that state in CI, and then each tests' job can rely on it; I'd wager the majority of requests in tests are for the same block, and this should help.
2 participants
No description provided.