Solana Proposes Removal of Durable Nonce Feature Following Drift Hack
Written with artificial intelligence.

Solana's developers are looking to eliminate the durable nonces feature, which was exploited in the $270 million Drift hack earlier this year. The proposal suggests a cut-off date of May 5, 2027, allowing time for custodians to adapt their systems.
Proposal to Remove Durable Nonces
Solana’s development team has put forward a draft proposal aimed at discontinuing the durable nonces feature, which allows signed transactions to remain valid for extended periods. This mechanism was exploited during the April hack of the trading protocol Drift, where attackers drained over $270 million.
Typically, transactions on Solana include a blockhash that expires after about 30 seconds. Durable nonces replace this blockhash with a permanent one, enabling transactions signed days or weeks in advance. This feature is particularly useful for custodians and exchanges that use offline (air-gapped) machines for approvals, as they often require longer processing times.
Timeline for Changes
The proposal, titled SIMD-0723 and submitted on November 5, 2026, specifies that the durable nonce feature will not be removed before May 5, 2027. This timeline gives users at least six months to adapt, starting from the proposal's publication date. However, the Solana Foundation's deprecation page indicates that the six-month period will only commence after a replacement is fully prepared, potentially delaying the end date further.
Impact on Users
While durable nonce transactions accounted for approximately 3% of Solana's network traffic in a recent measurement, they generated 30% of priority fees, indicating their significance in the ecosystem. A small number of accounts, specifically 555, were responsible for almost all durable nonce transactions. The proposal also highlights that custodians and exchanges using air-gapped signing would face the most significant adjustments.
Replacement Solutions
To replace the durable nonces, the proposal suggests the development of an on-chain program capable of verifying offline signatures. Options like Anza’s Programmatic Signer and Blueshift’s Vector are mentioned, but the proposal cautions that these alternatives come with their own risks, including the potential for failed executions to leave authorizations active, which could mislead custodians about transaction statuses.
This proposal awaits approval from key engineers at Anza and Firedancer before it can be implemented.
