X1 Labs has cleared Tachyon v3.1.14 for deployment on X1 mainnet, calling on the network’s validator operators to upgrade their nodes in a coordinated rollout announced today in the X1 Validators Army channel. The release is billed as a stability and refinement update that builds on the sweeping v3 performance overhaul shipped earlier this year — and validators who fail to upgrade within the allotted window will be forked out of the network sooner or later.

What’s in the release

The jump from v3.0 to v3.1 brings a series of upstream changes aimed at making nodes faster and more resilient. The headline item is a structural one: the accounts index is now kept fully in memory by default, with the on-disk index officially deprecated. The change speeds up account lookups across the board, but X1 Labs is explicitly warning operators to confirm their machines have sufficient RAM headroom before flipping the switch — a caution that proved warranted, as some validators reported having to update their kernel and raise memory limits before the new client would run cleanly.

Beyond the index change, the release hardens shred insertion and blockstore error handling, and cleans up the Turbine and gossip networking layers with IPv6 address filtering and CRDS table pruning. On the RPC side, getProgramAccounts now rejects malformed filters outright, and signature failures in simulateTransaction and preflight checks surface as a proper TransactionError rather than a generic JSON-RPC error — a small but meaningful improvement for dApp developers debugging failed transactions. The release also bumps several security-sensitive dependencies, including rustls-webpki, quinn-proto, and tar.

There is one X1-specific change of note for anyone watching the token’s numbers: circulating supply now excludes the X1 Foundation accounts, meaning getSupply queries and the circulating-supply figures shown on explorers finally reflect an accurate picture of XNT in public hands.

Deprecated flags

Three startup flags are deprecated in this release and should be removed from validator startup scripts:

  • --disable-accounts-disk-index — redundant now that the in-memory index is the default
  • --monitor (used with the exit command) — operators should run the separate monitor command after exit instead
  • --dev-halt-at-slot

New entrypoint keys

Separately from the release itself — and flagged as not urgent — X1 Labs announced it has rekeyed several entrypoint nodes. Operators are asked to update the known-validator set in their startup configuration to the following:

--known-validator 2skZy1HiefFby6a8raendfWGKGRmpcDiwvuPj1v7atch \
--known-validator 9zkHekQ7TS1JddcGMgMsZYDUc3xvGzUnXCAfKERvuwUR \
--known-validator ErccS37ii8NbPjsDdGso93vGj7rkF8WhuCV7GhoWmqsS \
--known-validator 8MUdMKKnod3zD1errE1jjeswLtuZ7jNiTvN2EC5cqbCG \
--known-validator 3ka4FFFTm3YgTJ17598ApHuzxELd7PcHnHQsY8uqbDsu \
--known-validator wwcXJz47MTnZacxJ4WSxbhtKuiZvHVqDrBbNCCpjVp1 \

How to upgrade manually

X1 Labs asks operators to hold their upgrade until network delinquent stake is below 5%, so restarts stay staggered and the network never loses consensus headroom, and to budget enough time to monitor the node after restarting. The client should be built from either the v3.1 branch or the release/v3.1.14 tag of the Tachyon repository:

cd ~/tachyon
git fetch --all --tags
git checkout v3.1        # or: git checkout release/v3.1.14
git pull
cargo build --release

From there the procedure follows the standard Tachyon upgrade path documented in the official X1 docs: stop the running validator, verify the new binary with tachyon-validator --version, restart with the updated startup script — including the new known-validator set — and confirm the node catches back up with solana catchup --our-localhost. Operators should then watch their logs and check solana validators to confirm the node is voting before walking away.

Upgrading through the X1 Console

For the large share of operators running the X1 Console, BlackBeard — the tool’s creator — published a parallel set of instructions. The critical detail: the console’s tracked branch must be changed to v3.1 first, or the auto-update will not work.

  • Step 1: Run Change Branch (option 10 → 2 → 3) and enter the new branch name: v3.1
  • Step 2: Run Update X1 Validator (option 10 → 2 → 1), after confirming delinquent stake is below 5%
  • Step 3: At any convenient time, run Update X1 Console (option 10 → 2 → 2) — a separate console update has been pushed to git carrying the new known-validator list

Community reports during the rollout suggest the path was not friction-free for everyone: some validators had to update the X1 Console itself before the branch change would take, and others — as noted above — needed kernel updates and increased memory limits to accommodate the in-memory accounts index.

The clock is ticking

The takeaway for operators is straightforward. The upgrade is not optional, and nodes left on older versions will eventually be forked out of the network. With the delinquent-stake gate keeping restarts staggered, the safest move is to upgrade early, watch the node closely after restart, and report any issues to the X1 Labs team through the validator channels.

The release is available now at github.com/x1-labs/tachyon. More on the X1 network at x1.xyz.