Maaz Zindani

Loading the index…

↑↓ move↵ openesc close4072 pages

Help for owners

I only find out my site is down when a customer tells me

Short answer

An uptime check requests your site from outside every minute or so and sends a message when it fails. A useful one tests a real page, the login and checkout, not only the home page, and sends alerts to more than one person. Google also reduces crawling of pages that keep returning server errors.

  • Written for business owners
  • Updated
  • By Maaz Zindani, Senior Software Engineer

Does this sound like your situation?

  • A customer or friend messages me that the site is not loading.
  • I discovered the outage hours after it started.
  • The server is fine on my phone but not for others.
  • Orders dropped and I did not know why.
  • I do not know who would be told if the site stopped.

Why does this happen?

  1. Nobody is watching from outside

    A site cannot report its own absence. A check has to come from a different machine, as a visitor would.

  2. A home page check misses the real failures

    The home page can load while checkout, login or the booking form is broken. The check must visit the pages that earn money.

  3. The alert goes to a single person or inbox

    One unread mailbox or one phone on silent means the first alert is missed.

  4. Outages hurt search visibility too

    Google states that pages persistently returning server errors are eventually removed from its index, and that crawling slows when errors appear.

What can I do about it today?

  1. Pick the three pages that matterFor example the home page, the login or booking page and the checkout or contact form.
  2. Choose a monitoring service with a free or low tierAny reputable uptime tool can check a page every few minutes. Read its alert options before choosing.
  3. Set alerts to two people and two channelsFor example email plus a phone message, so one miss does not hide an outage.
  4. Include the domain and certificate datesExpiring domains and certificates cause sudden outages that a monitor can warn about early.
  5. Write down who acts on an alertOne name, a phone number and what to check first.

When should I bring someone in?

  • The site has gone down more than once without warning.
  • The app has an API or background jobs that a simple page check cannot see.
  • You need alerts that reach someone technical.

How do I handle it?

  1. Decide what counts as downA failed login is an outage for a membership site, a failed checkout for a shop. I define checks from the business, not from the server.
  2. Add outside checks and inside checksOutside checks see what visitors see. Inside checks watch disk, memory, queues and scheduled jobs.
  3. Route alerts to a rotaMore than one person, each alert with a plain line saying what failed.
  4. Keep noise lowAlerts fire only when action is needed, so they are still trusted.
  5. Review after each incidentEvery outage adds a check or tightens one.

The price depends on what I find, so I quote in writing after a short discovery call. Tell me what is happening and I reply within 24 hours.

What do owners ask about this?

Is a free uptime checker enough?

For a simple site it is a good start. For checkout, logins or APIs it only helps if it tests those paths.

How often should it check?

Often enough that you would act on the gap. Shorter intervals find problems sooner but can create more false alarms.

Will monitoring stop outages?

No. It shortens the time before someone knows, and the fix still has to be made.

Sources

How was this page made?

I drafted this page with AI assistance from the official documents listed under Sources. Found a mistake? Tell me.

For your developer: the technical fixes

Next step

Want this off your desk?

Send me a few lines on where things stand. I reply within 24 hours.