Ecosystem v6.4.2

Release Date: August 13, 2026 Tags: EXPLORERS, CHAINS, DIAGNOSTICS, BUG-FIXES

Overview

A diagnostics release for the transaction-history providers. A working key rejected by its own provider's security settings reported nothing an operator could act on — the Blockchains → Requirements page showed the provider red with a bare HTTP 401, or with a rule name like Origin not allowed and no hint of where that rule lives.

Ankr is the provider this bites, and it never names the code as the cause, because it isn't: an Ankr project can restrict requests by website or by IP, and this platform calls Ankr from your server. A website restriction can never be satisfied — server-side calls carry no browser origin — so the key looks broken while being perfectly valid. Both are settings in the Ankr dashboard, and the page now says so.

A second defect sat underneath: Ankr reports failures in two different response formats and only one was read, so a wrong or missing key produced a status code and no explanation at all.

If your Ankr provider shows red, read Upgrade Notes — the fix is in your Ankr account, not in .env.

Requires Core v6.6.5.

Update Instructions

pnpm updator

Restart the backend afterwards. No database change and nothing to run.

Nothing in this release changes which provider serves a chain, or what any of them returns. If your chains are green today they stay green.

Upgrade Notes

An Ankr key can be valid and still rejected, and only the dashboard can fix it

An Ankr project carries endpoint-security rules, and two of them lock this platform out no matter what you put in .env. Open the Requirements page for the affected chain and read the Ankr pill — it now names the rule and the change that clears it:

All four are fixed in the Ankr dashboard under project → Settings → Security. For a server install, clear the website restriction and restrict by your server's outbound IP instead — a domain allowlist is meant for browser applications and locks a backend out by design.

• Fixed a key rejected by a domain or IP rule being retried on every chain indefinitely. It is now recognised as a bad credential and quarantined across all chains for ten minutes, because a domain rule rejects every chain identically. A chain missing from the project is treated separately, so a key that serves your other chains is not quarantined along with it.
• If Ankr is the only keyed provider on a chain, check that chain's pill row after fixing the dashboard — BSC, Fantom, Cronos, HECO and Polygon Amoy have no keyless fallback, so on those an unusable Ankr means no transaction history at all.

Ankr's endpoints no longer answer without a key

rpc.ankr.com/{chain} used to serve public traffic unauthenticated and now refuses every call — while still replying 200 OK, with the refusal in the response body. Any {CHAIN}_{NETWORK}_RPC still pointing at a bare Ankr URL is a chain with no node behind it: deposits stop confirming and withdrawals stop broadcasting, and older builds reported it only as a failed check.

• Search your .env for rpc.ankr.com in any _RPC or _RPC_WSS value and replace it with an endpoint that answers, or with an Ankr URL that carries your key. Run the chain's test afterwards.
• Check for a variable defined twice. A repeated key does not merge or warn — the last occurrence in the file wins silently, so a good endpoint on an earlier line is overridden by a broken one further down.

Changed

Blockchains → Requirements

• Changed the Ankr provider card to name the dashboard setting that clears each rejection, instead of repeating the rule Ankr broke on.

Fixed

Transaction-history providers

• Fixed a provider rejection reporting only a status code. Ankr returns failures in two different response formats and only one was read, so a wrong, missing or restricted key produced HTTP 401 or HTTP 403 with no cause named.
• Fixed an Ankr domain or IP restriction being classified as an unknown error, which retried it on every chain every thirty seconds instead of recognising the credential as unusable.