ViablyViably
Uptime monitoring

Know your site is down before a customer tells you.

Viably checks any public URL from outside your network, opens one incident in Slack when it stops answering, and closes it on the same message when it comes back. Nothing to install and nothing to deploy. Add the URL, pick a channel, and the board fills in from there.

Start freeSee pricing

Free plan forever with no credit card, and unlimited seats on every plan.

1 endpoint on the Free plan, 5 on Viably+.

#ops — Slack
Today
Viably square
ViaApp3:04 AM
🚨Checkout API is down

HTTP 504

Checked
https://api.athlead.com/health
Status
504

Two missed probes open the incident in the channel you chose. One miss is a deploy, so it says nothing.

The cadence

One miss is a deploy. Two is an incident.

Most monitors either page you on every blip or check so rarely that a five-minute outage reads as an hour. Viably runs three clocks: hourly while an endpoint is healthy, two minutes after a miss to confirm it, and every quarter hour while it stays down.

  1. 10:00
  2. 11:00
  3. 12:00
  4. 12:02
  5. 12:15
  6. 12:30
  7. 12:45

Healthy. Asked once an hour, from outside your network.

One outage, one message

An alert somebody will still read at 3am.

A channel that fills with the same alarm every two minutes is a channel people mute, and then the real one is missed. Viably posts once, edits that message as the outage runs, and writes the recovery on it when the endpoint answers again.

  • Two misses, then a card

    One failed probe is usually a deploy. Viably checks again two minutes later, and a second miss opens the incident in your channel.

  • Edited, never repeated

    The alert updates in place while the outage runs, with a reminder in its thread every six hours. On recovery, the same message shows how long it was down.

  • Or no channel at all

    Leave the channel empty and the check still runs, keeps its history and shows on the board. Useful for anything worth watching quietly.

The board

A week per row, six hours per cell.

Every endpoint you watch carries the last seven days beside its uptime and its median response time. A service that has been flapping since Tuesday shows up here before it becomes a complaint, and the probe that failed says why.

Uptime — Athlead
60 min
Between probes while healthy
15 min
Between probes while down
30 days
Of probe history kept

It reports itself on Monday.

Nobody opens a monitoring board to learn that nothing broke. With the power on, the Weekly Pulse carries its own uptime line: the endpoints that are down now, the ones that had an incident this week, and a single figure for the rest. The button under it opens the day-by-day breakdown in the thread, in your workspace timezone and workweek order.

  • The channel you chose

    One card per outage, opened on the second miss and closed on recovery. Reminders go in its thread to keep the channel tidy.

  • The Team Inbox

    A down endpoint sits in the inbox next to overdue tasks and customers waiting on a reply. Claim it, file it as an issue or hand it to someone.

#general — Slack
Viably square
ViaAppMon 9:00 AM

Weekly Pulse · Athlead

🚦 Uptime · 98.5% across 3 endpoints · 1 paused

Checkout API had 1 incident, 6h 41m on Friday. The other two were up all week.

🚦 Uptime by day
Viably square
ViaApp9:00 AM

Uptime, last 7 days · 98.5% across 168 hours watched

  • Mon100%
  • Tue100%
  • Wed99.6%
  • Thu100%
  • Fri93.3%
  • Sat100%
  • Sun100%
What counts as up

The rules, so the card is never a surprise.

  • A 200 is up

    Any endpoint that answers 200 within eight seconds is up, whether it’s an API, a marketing site, a webhook receiver or a status page.

  • JSON can report a failure

    A 200 whose JSON body says {"ok": false} counts as down. Many services report an internal failure that way.

  • Everything else is down

    A 5xx, a failed redirect, a timeout or an empty response all count as a miss, and the card names which one it was.

  • Public URLs only

    Probes run from outside your network, the same path your customers take. Private addresses and internal hostnames are refused.

Priced per workspace, and a probe costs almost nothing.

A healthy endpoint is asked twenty-four times a day, and each probe meters at a tenth of a message, so watching your site is a rounding error on the plan you already have. The Free plan watches one endpoint with the full feature set, including the board and the Pulse line, and Viably+ watches five.

1
Endpoint on the Free plan
5
Endpoints on Viably+
Seats on every plan
$7
per day, billed annually
FAQ

Questions about uptime monitoring

How quickly will I know my site is down?
A healthy endpoint is checked once an hour. The moment one check misses it goes back on a two-minute clock, and two misses in a row open the incident, so you hear within the hour and usually a good deal sooner. While it stays down the endpoint is asked every fifteen minutes, which is the accuracy of the outage length on the recovery card.
Why not alert on the first failed probe?
Because a deploy is a single failed probe, and so is a restart, a cold start and a brief network hiccup. An alert that fires on every one of those is an alert people mute. Viably counts the first miss, checks again two minutes later, and only opens an incident when the second miss confirms it.
Will a long outage keep pinging the channel?
No. The original alert is edited in place while the outage lasts, and a reminder is posted in its thread every six hours. When the endpoint comes back, the recovery is written on the same message. One outage is one message in the channel however long it runs.
What does a health endpoint have to return?
A 200 response within eight seconds. If it returns JSON, the body cannot say {"ok": false}. A 5xx, a failed redirect, a timeout and an empty response all count as down, and the error is recorded so the card can say which one it was.
Can I watch something that is not public?
Only if it is reachable from the public internet. The probes run from Viably's own infrastructure, outside your network, because that is the view your customers have. Private addresses and internal hostnames are refused when you add them.
Does every check have to alert in Slack?
No. Leave the channel empty and the check runs, keeps its history and shows on the board without opening incidents anywhere. You can add a channel later, or take one away, without losing the history.
What does the board show?
Each endpoint carries a week of six-hour cells beside its seven-day uptime and its median response time, so a service that has been flapping since Tuesday is visible before it becomes a headline. Every probe is kept for thirty days.
How much does it cost?
The Free plan watches 1 endpoint with the full feature set, and Viably+ watches 5. Viably+ is seven dollars a day billed annually with unlimited seats, and it opens with a 14-day trial that bills nothing until the fortnight is up.
Read the full guideUptime Monitoring in Slack
Ready when you are

Add the URL your customers hit. That is the whole setup.

Start free, paste a health endpoint, pick a channel. The first probe runs within two minutes and the board fills in from there.

Start freeSee pricing
SignalRecordOwned work
  • Free to start
  • No credit card required
  • Powers on when you need them