Network Reliability Watch / Network & Protocol

The official Teranode project has released version 0.15.8, introducing a recovery process for testnet and TeraTestNet nodes that stopped advancing because of an earlier UTXO-set classification error.

The underlying correction first arrived in Teranode v0.15.7 on July 25. According to the project’s release notes, the node software had been applying a pre-Genesis rule to outputs created after the Genesis upgrade.

Under the incorrect behavior, a locking script beginning with a bare OP_RETURN was treated as permanently unspendable and omitted from the UTXO set. When a later block contained a transaction spending that output, the node could not locate the required outpoint and stopped processing additional blocks.

Version 0.15.7 makes the rule dependent on the height at which the output was created. For post-Genesis outputs, only OP_FALSE OP_RETURN is treated as provably unspendable, while a bare OP_RETURN output remains in the UTXO set and may later be spent. Outputs created under pre-Genesis rules continue to be classified according to the earlier protocol conditions.

Correcting the storage rule prevents newly synchronized nodes from reproducing the problem, but it does not automatically repair a node whose local state was already built with the earlier classification. Version 0.15.8 therefore introduces a new teranode-cli rewindblockchain command.

The command allows an operator to rewind the local chain state past the affected block and then resynchronize from that point with the corrected v0.15.7 behavior in place. The release includes a dry-run mode, post-rewind consistency checks, interactive confirmation, adjustable subtree-loading concurrency, and explicit controls for unusually deep or non-idle rewinds.

Additional safeguards are intended to prevent the recovery process from accidentally advancing the recorded block-persistence height rather than moving it backward. The command also reads the node’s configured subtree-store hash prefix so that cleanup operations address the same storage paths used by the running node.

The release notes describe the rewind operation as destructive and include warnings around forcing the process while the node is not idle or rewinding more than 100 blocks. These options are intended for operators who understand the state and coinbase-maturity implications rather than as ordinary node-management actions.

The reported regression affected Teranode’s test networks. The release documentation does not describe a corresponding BSV Blockchain mainnet interruption. The incident and recovery path should therefore be understood as test-network operational hardening rather than a mainnet outage.

BSV TIMES read:
Test networks exist partly to expose state, validation, and recovery problems before infrastructure is relied upon more broadly. The important development is not only that Teranode corrected the UTXO rule, but that it added a defined recovery process for nodes whose stored state had already been affected. Reliable infrastructure requires both prevention and repair: correct protocol-era handling for future blocks, and controlled operational tools for restoring nodes when earlier behavior has left them unable to advance.

Teranode v0.15.8 release:
https://github.com/bsv-blockchain/teranode/releases/tag/v0.15.8

Teranode v0.15.7 release:
https://github.com/bsv-blockchain/teranode/releases/tag/v0.15.7

Posted on July 28, 2026