Repository navigation
Recover failed Axelar confirms from the funding account when the relayer fails - #1407
Merged
Merged
Conversation
Axelar's public signing relayer keeps signing ConfirmGatewayTx with a stale account sequence (Sep 1-2, Sep 29, Oct 6), so every recovery broadcast is rejected with code 32. ConfirmGatewayTxsRequest is permissionless, so a small AXL-funded account of ours can restart the poll without waiting on Axelar support.
✅ Deploy Preview for vortexfi canceled.
|
✅ Deploy Preview for vortex-sandbox ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
✅ Deploy Preview for vrtx-dashboard canceled.
|
…he relayer fails Axelar's signing relayer signed with a stale sequence again on 2026-10-06, so the automatic recovery could not restart the failed poll; a manual confirm from our own Axelar account recovered it. The backend now does that itself with the EVM funding key, which also controls an axelar1 account. The relayer stays first so environments without AXL keep today's behavior, and the manual script is gone.
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Why
recoverAxelarStuckConfirmdepends on Axelar's public signing relayer. The relayer keeps signingConfirmGatewayTxone account sequence behind the chain: 53011 vs 53012 in September, and 53019 vs 53020 on 2026-10-06. Every broadcast is rejected with code 32, and the Axelarscan "confirm" button fails the same way.ConfirmGatewayTxsRequestisROLE_UNRESTRICTEDin axelar-core v1.5.3, so any funded Axelar account can restart the poll. On 2026-10-06, a confirm sent from our own Axelar account recovered 0x9813…8ff7: Axelar txEAA8FE9D…87E3, after which the call was approved and executed. This PR makes the backend do that by itself.What changes
maybeRecoverStuckConfirmtries the relayer first, as before. If the relayer fails (except on abort), the same attempt falls back toconfirmGatewayTxFromFundingAccount.confirmGatewayTxFromFundingAccount(squid-router-swap/axelar-confirm.ts) signs withEVM_FUNDING_PRIVATE_KEY. That secp256k1 key also controls anaxelar1…account. The function:axelar1…address to fund, which shows up in the stuck-GMP Slack alert;axelarConfirmRecoveryAtcooldown covers both attempts, so spending is capped at one confirm per stuck ramp per cooldown.@cosmjs/proto-signing,@cosmjs/stargateandcosmjs-typesare added, inapps/apionly.Before deploying
Send a few AXL, on the native Axelar network, to the funding key's
axelar1…address. It differs from the EVM address0xEC733…33e2. If it's unfunded, the first fallback attempt fails and the alert names the address.Tests
A2D28977…41B6(code 0).confirmGatewayTxFromFundingAccountwith a mocked client:The two new executor tests fail without the change.
bun test squid axelar(api): 40 pass, 0 fail, 2 skipped;tscclean.