FAQ

Short answers, with links to the full explanation.

Setting up

Do I install anything?

No. Roundbeat runs its own beacon node and reads Lido’s registry and contracts. Find your operator, press Monitor, then add a destination and a rule (Getting started).

Is it free?

Yes, today. Every dashboard is public, and monitors and alerts need only a sign-in. Billing is not enabled yet, so nothing is charged; the Billing tab shows the plans that will come. An organization can have up to 10 monitors.

Which networks and modules?

Ethereum mainnet, with every operator in Lido’s Curated v1, SimpleDVT, CSM and Curated v2 modules, including the move from Curated v1 to Curated v2.

Who can see my monitors?

Only your organization’s members. Operator pages are public, but which operators you monitor, your rules, destinations and alert history are not. A withdrawal-address monitor’s page is visible to your organization only.

I monitor my Curated v1 operator. Do I get alerts for its Curated v2 operators?

No: monitor the Curated v2 operators directly, since each is a separate operator in its own module. Your Curated v1 monitor keeps the migration alerts about the Curated v1 side (migration toggles).

Alerts

How soon do I hear about an offline validator?

No sooner than about 19 minutes after a node goes down (see Offline alerts are not instant).

An event shows on the page, but I got no alert. Why?

Check, in order:

• a rule covers the monitor and lists the type (or All events);
• the rule’s minimum severity lets it through, and its status is Active;
• the destination is Verified and is not email;
• the same alert was not sent within the rule’s cooldown;
• the event is not one that never alerts, and not too old to alert (late events).

Why did I not get a recovery message?

A recovery goes only to destinations that received one of the outage’s alerts, and only after every key is back and 30 minutes pass with no key of the operator going offline. A key that goes offline in the meantime keeps the outage open (offline outages).

PagerDuty did not page for a second wave of offline keys.

By design: keys going offline within 2 h of the outage’s last alert belong to the same outage and update its open incident. A wave after that opens a new incident that pages.

My rule says “Never fires”.

Its minimum severity is above every type it lists, for example Critical on Proposal hit, which is info. Lower the minimum or add a type that reaches it.

Why does an alert start with “⏱ Delayed”?

Roundbeat caught up after falling behind, and the event was older than 2 h but important enough to send anyway (late events). The body says when it happened.

Can I get alerts by email?

Not yet. Use Telegram, Slack, PagerDuty or a webhook.

Why does the upcoming-proposal alert come so late?

The beacon chain publishes proposers only one epoch ahead, so Roundbeat can announce a proposal at most about 13 minutes before its slot.

Data

Why does a number show “—” instead of 0?

Roundbeat does not know it: the data is missing or not indexed yet. A real zero shows as 0.

My Curated v1 validator left the active set but is not “Exited”.

It is consolidating into Curated v2, which is not an exit. It shows as Consolidating → CMv2, then Consolidated → CMv2 (see The Curated v1 to v2 migration).

Why is a value faded?

Too little data to judge: effectiveness over fewer than 50 epochs, or a proposal rate over fewer than 10 proposals.

Why are depositable keys missing for CSM?

CSM does not report depositable keys per operator, so Roundbeat shows nothing rather than a guess, and raises no low-depositable warnings for CSM.

Why is a block’s relay data still pending?

A relay has not answered yet. When a relay answers “too many requests”, Roundbeat leaves it alone until its cooldown ends, and the block waits for the next check. Checks continue for 24 h after the slot (see How long a block can stay pending).

What does “Chain data delayed” mean?

Roundbeat’s beacon node is more than 64 slots behind the chain, and the pages show the last data they have. The banner sums up the effect on alerts: “Beacon-chain alerts lag with it (most are skipped once 2 h late), exit-deadline reminders wait, and contract alerts are not affected.” In more detail:

• Beacon-chain events arrive only as fast as the node does, so they and their alerts run the same lag behind. A stuck node produces none.
• Once the node catches up, an event older than 2 h alerts only if it is still worth it (late events).
• Exit-deadline reminders wait while a validator’s beacon state is more than 30 minutes old.
• On-chain contract alerts come from the execution chain and keep coming.

The API reports the same state at /api/status.

API and sharing

Do I need an API key?

No. The public API is read-only and open. It is meant for scripts and servers: other websites cannot call it from a browser.

How do I point a colleague at one missed slot?

Highlight its row and copy the link: the slot travels in the URL as hl=slot.<slot>, up to 20 items per link (see Shared links and highlights).

How do I share an operator’s missed proposals for the last week?

On the operator’s Proposals tab, filter to missed, pick 7d and press Share. The link carries status=missed and window=7d, and its preview states the same count the tab shows (see Time window on Proposals and Events).