Frequently asked questions

Many provider signals wait for several bad rounds in a row, 2 by default. A single miss, for example during a restart, stays quiet on purpose. If you want to hear about every miss, set Consecutive rounds to 1 for that signal. More in Consecutive rounds.

The problem lasted exactly as many rounds as needed for the alert, and the next round was fine again. This is typical for short hiccups such as a restart. If such pairs bother you, raise Consecutive rounds or switch off the recovery notice of that signal. See Recovery notices and cooldown.

Your signature for a voting round was not submitted within the grace period that counts for rewards. It arrived later or not at all. This alert comes immediately, without waiting for more rounds. Check your signer and its timing; details are in Alerts without waiting.

A signature alert adds one explorer link for every phase that was actually affected. If the FTSO and the FDC signature both missed, you see two links; if only one missed, you see one. Open each link to see which transaction came too late. More in Alerts without waiting.

Validator not listed only fires when none of your registered NodeIDs is a current validator. As long as one validator is still listed, for example while you reinvest the other one, there is no alert. See Validator not listed.

During quiet hours, private Telegram chats, Discord direct messages and e-mails are held. After the window ends, each of these recipients gets one summary. Telegram groups, webhooks and the bell keep delivering, so route urgent alerts there. See Quiet hours.

First check that Send recovery notice is on for that signal. Wallet balance changes and Feed price moves have no recovery notice at all. For the others, the Watchdog must see proof that everything is fine again; missing data does not count. See Missing data is never "all good".

Routing decides who gets which alert: a signal group can have its own recipients, and a watch can use its own selection instead of the defaults. Recipients that are still waiting for verification, or that are unreachable, get nothing, and quiet hours hold personal channels. Check section 4. Recipients of the watch and Default recipients and routing.

In the Alert history. Set the filter Delivery to Failed only, then hover over the chip of a recipient to see the error. See Searching and filtering.

The Watchdog tries each delivery up to 3 times in total. After that it stays failed in the history. If the target refuses permanently, for example because the bot was blocked, the recipient is marked Unreachable; see Unreachable — what to do?.

Webhooks must use public HTTPS on port 443 or a port from 1024 upwards; private and internal addresses are rejected. Discord and Slack webhooks become unreachable when they were deleted on their side. Fix the URL or create a new webhook, then press Send test. See Webhook.

For signal alerts, yes: switch off Alerts in the notification bell in Notification channels. Account and system messages, such as a suspended watch, always arrive in the bell. See Notification bell and system messages.

No. Watches on your own credit balance are free, up to a limit per wallet. The limit is listed in Costs and limits per type.

If auto-renew is on, the pack renews from your credits; if that fails, it keeps trying for 7 days. Without a slot, your newest paid watches are suspended until you buy slots and press Resume. See When slots expire.

No. A paused watch keeps its slot so that you can resume it at any time. To free the slot, delete the watch; its alerts stay in the history. See How slots work.

Open Account Settings → API access and press Revoke next to the token. Access with that token stops immediately. See Overview and revoking.