StakeSafe

The Eagle-Ray adoption tracker (10 September 2026)

Radix mainnet had produced no rounds since 21:19:48.939 UTC on 31 August 2026, when validators holding more than two thirds of stake broke liveness deliberately after every Hyperlane-bridged asset on the network was drained. It restarted on 11 September; the three sections that follow were written while it was down, and each is a dated reading. The fix shipped as babylon-node v1.4.0.0, “Eagle Ray” on 10 September, and at 16:15 UTC that day the Radix Accountability Council told node operators to install it. No restart date has been published. What has been published is a threshold.

At 20:18 UTC on 10 September 2026, in the main Radix Telegram group, StakeSafe's Bart Roozeboom announced that the operator's free Radix Network Dashboard now tracks Eagle-Ray adoption live: the share of active validator-set stake already running v1.4.0.0 against the share still on older versions, as a chart and a breakdown table. The announcement states the condition in one line – Once more than 67% of active stake is on Eagle-Ray, network liveness resumes and the network forks to a patched version – and the dashboard repeats it in a banner above the charts. The message is authorship-verified as Bart Roozeboom at its public embed. The 67% is not a StakeSafe rule: it is the ordinary two-thirds quorum Radix's consensus needs to commit a round, which is the same reason the network could be stopped by the same fraction of stake in the first place.

Read from the dashboard's served page at 23:08 UTC on 10 September 2026, seven hours after operators were cleared to upgrade:

Two details in that table are worth stating precisely, because the dashboard does not. The percentages are computed on 4,663,021,341 XRD – the active validator set, the top 100 by stake, which is the dashboard's own online and offline figures added together – and not on the 4,877,281,907 XRD the same panel heads with as total delegated stake; the two differ by 214,260,566 XRD, so a figure quoted against the wrong base runs about 4.4% high. And the upgraded stake is concentrated: 28 rows in the explorer table read v1.4.0.0, and their delegated stake sums to the 1,260,200,692 XRD the panel rounds to 1,260,200,690.

The gap is forty percentage points, and most of it sits behind nodes that are switched off rather than behind nodes running the wrong version: two thirds of the active set's stake is offline. Restarting therefore asks operators to come back and to come back upgraded, which is why the council's instruction is to leave a fully upgraded node online and working rather than merely to install the release. The dashboard refreshes itself every five minutes, so any reading of it – including this one – is a snapshot rather than a settled figure.

Where the gap sits (11 September 2026)

Read from the dashboard at 07:12 UTC on 11 September 2026, eight hours after the reading above, both panel figures had risen: 1,455,804,324 XRD is running v1.4.0.0, 31.22% of the active set, against 1,260,200,690 and 27.03% the previous evening; stake with a node online rose further over the same hours, from 1,519,267,322 and 32.58% to 1,746,449,780 and 37.45%. The two lines are separating rather than converging, because operators are coming back online faster than they are coming back upgraded: the stake that is online but not on Eagle Ray widened from 5.6 points of the active set to 6.2.

The explorer table below the panel says where the rest of it sits. Thirty-nine rows read v1.4.0.0 and their delegated stake sums to 1,455,804,327 XRD, reproducing the panel's adoption figure; thirty-eight of those thirty-nine have a node online. None of the twelve largest validators is among them. Ranked by delegated stake those twelve are SRWA, Astrolescent, Reddicks, Weft, Jazzer9F, Ocinode, DeFiPlaza's investment node, Radical Staking, Avaunt Staking, both CaviarNine validators and RadixStake; every one reads a v1.3 version and every one reads offline. Together they hold 2,073,397,316 XRD. The shortfall between current adoption and the 3,124,224,299 XRD that more than 67% of the active set comes to is 1,668,419,975 XRD, so those twelve operators hold between them more than enough to end the halt, and the largest validator that has reached Eagle Ray, XSEED Staking at 98,684,205 XRD, ranks thirteenth.

One caveat on the version column, because it changes what the table proves. The explorer reports the version a node last advertised, so an offline validator shows the version it was running when it stopped. A v1.3 reading against an offline node means the explorer has not seen that node on Eagle Ray, not that its operator has not installed the release; the council's instruction is to install it and leave the node online, and only the second half of that is visible here. What the table establishes is where the stake is, and that the threshold cannot be reached by the operators already running. What it does not establish is why those twelve are down, and the answer was published the same morning: it is deliberate.

What happens once it is reached is fixed in the node software rather than announced: the fork enacts at the start of epoch 339,898 without a readiness vote, after an epoch in which rounds are produced but user transactions are refused. The sequence, and the reading of the halt that goes with it, are on Radix Ecosystem Operational Status.

Offline by design (11 September 2026)

Read from the dashboard at 11:05 UTC on 11 September 2026, three hours and fifty-three minutes after the reading above, 1,528,203,682 XRD is running v1.4.0.0, 32.77% of the active set, and 1,848,850,528 XRD, 39.65%, has a node online. Adoption rose by 72,399,358 XRD and 1.55 points over those four hours and none of it came from the top of the table: the same twelve rows, in the same order, still read a v1.3 version, still read offline, and still hold 2,073,397,318 XRD between them. Forty-one rows now read v1.4.0.0 against thirty-nine at 07:12, and the shortfall to the 3,124,224,298 XRD that more than 67% of the active set comes to has narrowed to 1,596,020,616 XRD. The whole of the movement is in the tail.

The reason the twelve have not moved was stated in the main Radix Telegram group two hours before that reading, and it inverts what the table appears to show. Asked why the ten largest validators could not spare ten minutes to update, Faraz replied at 10:11:24 UTC that the state is intentional: The largest nodes are upgraded and waiting to boot up together once we have a decent amount of stake online from the remaining nodes. The aim is to be well clear of 67% rather than barely over it, because crossing the threshold marginally risks a node developing a problem and falling over, triggering another liveness break, and a handful of large nodes is easier to coordinate than a long tail of small ones. Daffy had answered the same complaint sixteen minutes earlier in fewer words: the update is planned this way, and any conclusion about a validator still on v1.3 should wait until a week after the network is live. Both messages are authorship-verified at their public embeds (Faraz, Daffy).

Two things follow for anyone watching the chart. The caveat recorded here at 07:12 UTC was the operative one rather than a hedge: a v1.3 reading against an offline node says only that the explorer has not seen that node on Eagle Ray, and for the twelve largest it is now sourced that the release is installed. And the restart will arrive as a step rather than a climb. The stake that closes the gap is held by nodes that are down on purpose, so adoption can sit well short of 67% until the moment it clears, and the chart measures the tail's progress rather than counting down to a restart.

One class of row is not covered by that explanation, and it is the class the version column reads without ambiguity, because the staleness the caveat describes applies only where the node is down. Eighteen validators read online and read a v1.3 version, holding 327,184,952 XRD, 7.02% of the active set: nodes that are up and advertising a build the threshold does not count. The two largest are Cadwynbloc at 64,436,221 XRD and Supreme Stake at 51,025,921 XRD, and the set also holds the Community Council Node, HUG, Blockshard, Allnodes and DeXter. These are the operators the council's instruction is aimed at, and they are also why the two panels cannot be read as one: the 6.88-point spread between online stake and Eagle-Ray stake is this class, less the single mirror row, Atlas-Staking.com, which reads v1.4.0.0 and offline at 6,538,106 XRD.

The ledger was unchanged by any of it. Read at 11:05:08 UTC on 11 September, gateway-status returned state version 557,840,622 at epoch 339,896, round 102, and reported itself the same interval behind the ledger. That is the aggregator’s position rather than the ledger’s: the last state committed before the stop was 557,840,627, epoch 339,897 round 4, five states and forty-two seconds further on, and the status endpoint never showed the difference. Thirty minutes after this reading the network was running again.

How it resolved (11 September 2026)

The tracker above stops half an hour short of the thing it was tracking. Radix mainnet certified its first round in ten days at 11:35:28.96 UTC on 11 September 2026 – state version 557,840,628, epoch 339,897 round 5 – thirty minutes after the reading above. Epoch 339,897 then ran to its end producing rounds and refusing user transactions, and at 11:39:25.129 UTC epoch 339,898 opened at round 2 with the Eagle Ray fork enacted and the moratorium lifted together. Read from the transaction stream, the boundary is three entries: 557,840,627 at 21:19:48.939 UTC on 31 August, 557,840,628 at 11:35:28.96 UTC on 11 September, and sixty-five further round changes before the first user transaction at 557,840,694. The ledger produced no round for ten days, fourteen hours and sixteen minutes.

The reading this page took from the sequence held. Adoption did arrive as a step rather than a climb: it stood at 32.77% of the active set at 11:05 UTC, and the fork enacted with 80.85% of active-set stake on the release, so the twelve largest validators booted together as Faraz had said they would and the chart never counted down to anything. The dashboard's value was in showing where the stake sat rather than in predicting the hour, and on that it was accurate throughout. The full sequence and its ledger record are on Radix Ecosystem Operational Status.

The fleet after the Avaunt handover (29 August 2026)

Read live from the Radix Gateway validator list at epoch 339,152 (29 August 2026, 07:13 UTC), the handover announced for 28 August has completed, and the evidence is custody rather than announcement. The Avaunt Staking validator's owner is now the account account_rdx16xrp5e8…kpkdl84f – the same account that holds StakeSafe Amsterdam's owner badge. Rotterdam and the seed node sit in two further accounts, so the operator does not run its fleet from one key, but the acquired node and Amsterdam share one.

The second change answers the question this page could not answer on 21 August. The incoming validator's fee has now been stated, and it was stated on ledger: a validator_fee_change_request raising its fee factor from 0.02 to 0.25, effective at epoch 342,482. With that request the whole fleet converges on the same number.

That is 299,581,425 XRD across four validators, or 6.08% of the 4,928,844,882 XRD staked to the network's 287 registered validators at the same reading – up from the 157.1 million XRD across two nodes this page recorded eight days earlier. The dates are estimates derived from the five-minute target epoch length; the epoch numbers are the commitment.

Two of those repricings land within about a day and a half of this reading. None of them is visible on the validators themselves: all four still display the fee they charge today, because a queued change lives in the fee-change request rather than in the fee the validator reports, and a delegator reading the name, the fee and the website sees nothing at all. It is the same shape as the handover itself, where the sale was invisible in every field the validator published and legible only in where the owner badge sat.

Ledger position and the queued fee increase (August 2026)

Read from the Radix Gateway at epoch 336,990 (21 August 2026, 19:05 UTC), both validators are registered and accepting delegated stake. StakeSafe Rotterdam holds 81,182,620 XRD at rank 21 and StakeSafe Amsterdam 75,944,786 XRD at rank 23, of 188 registered validators and 4.69 billion XRD staked to them — 157.1 million XRD combined, about 3.4% of all staked XRD. The two nodes have swapped places since this page was last written: Amsterdam was the larger of the pair in July 2026.

Both validators carry a pending fee change already recorded on ledger. The validator_fee_change_request on each raises the fee factor from the current 0.15 to 0.25 — from a 15% to a 25% validator fee — effective at epoch 339,608 for Amsterdam and epoch 339,609 for Rotterdam. Radix's mainnet genesis config targets 300,000 ms — five minutes — per epoch, which puts those roughly nine days after the reading above, or approximately 30 August 2026; the epoch is the exact commitment and the date is an estimate derived from it.

The delay is not the operator's choice but the protocol's. The same config sets num_fee_increase_delay_epochs to 4,032 epochs — about two weeks — and the Radix Engine's validator blueprint applies it only when the new fee factor is higher than the current one; a fee cut takes effect immediately. Working backwards from the effective epochs, the two requests were signed at about epoch 335,576, on or around 16 August 2026 — before the Avaunt announcement, not in response to it. A pending request is not final either: a further update_fee call replaces it, so the owner can still lower or cancel the increase before it lands, and delegators have the two-week window to move if they would rather not pay it.

There is a third request, on a third validator, and it is a different kind of number. StakeSafe Seed Node — registered, accepting delegations and holding a token 210,099.38 XRD — carries a validator_fee_change_request with a new fee factor of 1: a 100% validator fee, also at epoch 339,609. A fee that takes every emission is not a price, it is a way of asking delegators to leave, and it is the same move Leaf Node made on 10 August 2026 eleven minutes before it deregistered. StakeSafe has announced nothing about the seed node's future, so this is recorded as what the ledger shows rather than as a stated plan; the two production validators are unaffected, and this page's other numbers were read in the same pass over the full mainnet validator list at epoch 336,991.

Why the fee matters this month

The rise is scheduled to take effect within days of StakeSafe taking on Avaunt Staking's node, announced on 21 August 2026 for 28 August. Avaunt has been running its validator at a 2% fee; delegators who follow the operator's advice and leave their stake in place will be staking under a materially different arrangement. The two facts are independent — the fee requests were signed around 16 August, before the Avaunt announcement, and apply to StakeSafe's own two nodes rather than to the incoming one — but they land in the same week, and the incoming validator's fee after the handover has not been stated.

Free tooling run for the network

Alongside the validators, StakeSafe publishes infrastructure it does not charge for: the Radix Network Dashboard (live, serving a full validator-set view), StakeSafe Alert for node runners and delegators, and a seed node and an outgoing node for the peer-to-peer network. The operator states it has run Radix nodes on its own hardware since the Olympia betanet days, across multiple regions and fully non-custodial.

StakeSafe is a product that is designed to provide high-performance nodes, secure staking, and professional support for the decentralized Radix ecosystem. The product's focus is on ensuring the efficient and secure operation of the Radix network through the provision of reliable nodes that enable secure staking and provide a dedicated support team that can help users with any issues that they might encounter.

By using StakeSafe, users on the Radix network can benefit from reliable and high-performance nodes that provide them with fast transaction processing speeds and low latency. Additionally, the product also enables users to securely stake their tokens while maintaining control over their assets, providing them with greater security and peace of mind.

Overall, StakeSafe is a product that aims to empower users and contribute to the growth and success of the Radix ecosystem by providing cutting-edge infrastructure and support that enables users to transact on the network with confidence and ease.

Mission

The mission of StakeSafe is to empower and contribute to the decentralized Radix ecosystem by providing high-quality infrastructure, secure staking services, and professional support.

Benefits

Delegators can benefit from staking with StakeSafe for several reasons:

Enhanced Security

StakeSafe prioritizes secure staking, ensuring that delegators' assets are protected. By utilizing secure staking mechanisms, delegators can have peace of mind knowing that their tokens are being staked in a safe and reliable manner.

Reliable Infrastructure

StakeSafe provides high-performance node, which are essential for efficient and effective staking. Delegators can rely on StakeSafe's robust infrastructure to ensure smooth and uninterrupted staking operations.

Dedicated Professional Support

StakeSafe offers dedicated support from a team of professionals. This support can be invaluable for delegators who may have questions, need assistance, or encounter any issues during the staking process. The responsive and knowledgeable support team can provide timely help and ensure a positive staking experience.

Active Participation in the Radix Ecosystem

By staking with StakeSafe, delegators actively contribute to the decentralized Radix ecosystem. Their participation and stake help secure the network while potentially earning rewards in the form of incentives or staking rewards.

Source: StakeSafe – the operator runs its two production validators (Amsterdam & Rotterdam) plus a seed node on the Radix mainnet.

Impact

StakeSafe can have several impacts on users and the wider ecosystem. Here are a few potential impacts of using StakeSafe:

Staking Incentives

StakeSafe allows users to stake their tokens and participate in the network's consensus protocol. By doing so, users can potentially earn rewards or incentives for contributing to the network's security and operation. This can be a significant incentive for users to actively participate in the ecosystem.

Network Security

Staking services like StakeSafe play a crucial role in maintaining the security and integrity of the network. By staking their tokens, users contribute to the network's decentralization and consensus process, making it more resilient to attacks or malicious activities.

Ecosystem Growth

Delegators using StakeSafe contribute to the health and growth of the ecosystem. By actively participating in the network, users help ensure its stability and attract more participants, developers, and businesses to the ecosystem. This increased activity and engagement could lead to a flourishing Radix network.

User Convenience

StakeSafe aims to provide a user-friendly staking experience, simplifying the process for users who may not have the technical expertise or infrastructure to stake directly. This convenience can encourage more users to participate and stake their tokens, boosting overall network participation and security.