Glossary
Terms you will encounter in the COTI Node Ecosystem web app, the installer, and this documentation. Definitions are authored for node operators; wording may differ from internal field names while preserving the same meaning.
Node states
Node status
The operational state of the node process, derived from the node's JSON-RPC eth_syncing method and peer presence:
Active — the node is online and fully synced with the tip of the chain.
Syncing — the node is online but still catching up on blocks.
Offline — the node is not currently observed in the peer set.
"Online" in the UI refers to any node that is Active or Syncing.
NFT state
The on-chain thermal flag stored on the node's Soulbound NFT:
Hot — the NFT has been minted and is currently marked hot.
Cold — either the NFT has not been minted yet (treated as cold by default), or the NFT has been marked cold after a prolonged outage.
Thermal status
The combined state shown to operators, derived from node status + NFT state:
Active / Syncing
Cold
Warming up
Active / Syncing
Hot
Hot
Offline
Hot
Cooling down
Offline
Cold
Cold
Warming up
A node is warming up when it is online and reachable but has not yet been continuously present long enough for the ecosystem to consider it stable. No NFT has been minted yet. The operator should keep the node online — the warm-up progress bar in /my-nodes tracks remaining time.
Cooling down
A node is cooling down when it was previously hot (NFT minted) but has gone offline. If the node comes back online before all connected time in the rolling 103-hour window has aged out, it can remain hot. If it stays offline until peer presence in that window reaches zero, the NFT flips to cold and the node must warm up again (~72 hours of connected time within the window).
Time to thermal update
Time remaining until the node transitions thermal state — until it becomes hot if currently warming up, or until it becomes cold if currently cooling down. Not shown when the node is already in a stable state (hot-while-online or cold-while-offline).
Warm-up & cooldown windows
The thermal state machine is driven by four configuration values maintained by the ecosystem. They are exposed so operators understand the timing of their node's transitions.
HOT_WINDOW_HOURS
The length of the rolling window used to decide whether a node is stable enough to become hot. Defaults to 103 hours (one epoch).
HOT_THRESHOLD_HOURS
The minimum number of hours within HOT_WINDOW_HOURS that a node must be present as a peer to qualify as hot and receive its Soulbound NFT. Defaults to 72 hours.
In plain language: "To become hot, a node must be seen by peers for at least HOT_THRESHOLD_HOURS hours during the last HOT_WINDOW_HOURS hours."
Becoming hot — illustrated example (production values)
In production the ecosystem uses HOT_THRESHOLD_HOURS = 72 and HOT_WINDOW_HOURS = 103. Peer discovery evaluates a rolling window: at any moment, only the last 103 hours count. Connected time that started before that window has already slid out and no longer contributes.
How to read the diagram
A
Node came online recently and has been up steadily
50 h
Still warming up — 22 h short of 72
B
Same node, still online, a while later
72 h
Hot — threshold met inside the window
C
Node was connected for 80 h, then went offline; by the time it is evaluated again, the first 27 h of that streak are older than 103 h
53 h (80 − 27 lost)
Not hot — earlier connected hours are lost for warm-up purposes
The window moves forward continuously. Every hour that passes drops the oldest hour from the count and adds a new empty hour at the trailing edge — unless the node stays connected, in which case that new hour accrues as connected time.
COLD_WINDOW_HOURS
The length of the rolling window used to decide whether a previously-hot node should be demoted back to cold. Defaults to 103 hours (one epoch).
COLD_THRESHOLD_HOURS
Configured alongside COLD_WINDOW_HOURS; in production both default to 103 hours. Peer discovery marks a hot node cold when it has zero peer presence anywhere in the rolling COLD_WINDOW_HOURS window.
In plain language: "A hot node that stays offline until all connected time in the last 103 hours has aged out cools back to cold and must redo the warm-up (~103 hours of continuous absence)."
Identity & ownership
Node private key
The 64-hex secret that identifies a node on the network. It is generated locally in the browser (or supplied by the operator), written to the node's nodekey file by the installer, and never transmitted to any COTI server.
Node address (public key)
The Ethereum-style address derived from the node private key. It is used as the node's identity on-chain and is the wallet that receives the Soulbound Node NFT.
Wallet address
The wallet connected to the web app (for example via MetaMask). For the per-operator dashboard to show a node, the connected wallet must be the node address — that is, the wallet that owns the Soulbound NFT.
FQDN (Fully Qualified Domain Name)
The public hostname the operator uses for their node. It may be your own domain (for example node1.example.com) or a COTI-assigned name when using the tunnel installer (--with-frp), typically under a managed zone such as *.testnet.nodes.coti.network on testnet.
The FQDN is a prerequisite for rewards: the ecosystem probes JSON-RPC through that name to determine uptime. See Installation, Own domain, and Wizard tunnel.
RPC URL
The public HTTPS endpoint the node exposes after installation, typically https://<your-fqdn>/rpc. After your node becomes hot and a monitor is registered, Better Stack polls the Node Health Monitor, which in turn calls this URL to verify block progression — Better Stack does not call your node directly.
Soulbound Node NFT
A non-transferable NFT minted to the operator's wallet once the node becomes hot. It stores the node's name and image and is the on-chain proof of node ownership. Because it is soulbound, it cannot be transferred between wallets.
Rewards
Epoch
A fixed 103-hour reward period. At the end of each epoch, the rewards service evaluates eligibility and credits eligible nodes.
Eligibility
The set of rules a node + operator must satisfy in an epoch to receive rewards. The web app describes two paths; you qualify if either path is fully met (see /eligibility):
Path 1 — USDC + COTI. USDC on the COTI network ≥ combo threshold and COTI (non-custodial; not CEX) ≥ combo threshold and uptime ≥ configured percentage.
Path 2 — COTI only. COTI ≥ solo threshold (no USDC) and uptime ≥ configured percentage.
Whitelisted operators are exempt from the holdings rules and only need to meet the uptime rule.
Eligible
A node that was found eligible for rewards in at least one epoch. The home-page stats card counts unique nodes that have ever been eligible; the /my-nodes per-epoch history marks each epoch with Eligible / Ineligible.
Whitelisted
An operator flag that exempts the wallet from the Path 1 / Path 2 holdings checks. Uptime is still required. Whitelisting is managed centrally by COTI; most operators will not be whitelisted.
Blacklisted
An operator flag that prevents a node from being accepted as a peer candidate. Blacklisted nodes do not appear in the ecosystem's dataset and cannot accrue rewards. Blacklisting is used for operational or abuse reasons and is not a state a normal operator encounters.
Uptime
The percentage of an epoch during which the node was judged healthy by the monitoring stack. An RPC that merely answers is not enough — the health check confirms that the node is actually operating, and only healthy results accrue uptime. Displayed both per-epoch and all-time.
Rewards (total / last epoch)
Total Rewards — all-time COTI earned by a node across every epoch it was eligible for.
Last Epoch Rewards — COTI earned in the most recently closed epoch.
Next Epoch Rewards — the size of the pool that will be distributed in the upcoming epoch (configured centrally).
Rewards accrue in the on-chain rewards smart contract rather than being deposited directly into the operator's wallet. Operators withdraw their balance via the Claim Now button in the My Node dashboard, or by calling the rewards smart contract directly.
Other terms you may see
Live node heartbeats
The home-page visualization showing the count of nodes currently observed by peer discovery. The pulsing tiles are purely decorative; only the counter reflects real data.
Better Stack
The external uptime-monitoring platform used by the ecosystem to probe every hot node. Operators do not sign up for or configure Better Stack directly — the ecosystem registers monitors automatically after NFT minting.
Status page
The public Better Stack dashboard that aggregates every hot node's monitor state (up / down). It is the fastest way to see the current health of the whole fleet. URLs are listed in Networks.
Operator status dashboard (local)
A small local web UI shipped with coti-full-node, bound to localhost at port 8090 after ./start_coti-full-node.sh. It shows process health, peer count, sync state, and (when configured) DNS/TLS or FRPC gateway checks — distinct from the Nodes web app dashboard at /my-nodes. With host Nginx or the COTI tunnel (via the internal nginx-frpc-gateway container), the same page is also available at https://<fqdn>/operator/.
Last updated
Was this helpful?