Skip to content
Brainrune DocsDocs

Alerts

Runs can keep happening while pages come back cold. The share of warmed URLs that still answer MISS is the number to watch, and the warmer can email you when it gets too high.

Turn alerts on

Section titled “Turn alerts on”

  1. Open Cache Warmer > Settings, panel Alerts.

  2. Turn on Email alerts.

  3. In Send to, type the address, or leave it empty to use the site admin address.

  4. Set Miss threshold, the share of warmed URLs answering MISS that should trigger an email. The default is 40%.

  5. Click Save settings, then Send a test alert to check that your site can send mail at all.

Alerts go out the same way as every other WordPress notification. If the test never arrives, fix mail sending on the site first, for example with an SMTP plugin.

When an alert is sent

Section titled “When an alert is sent”

An alert is sent when all of these are true:

  • The cold share has been measured. That needs a second pass: on a first pass every URL answers MISS, so nothing is judged before then.
  • At least Minimum sample URLs were measured (default 50). Fewer than that is too little to tell anything.
  • The cold share is at or above your Miss threshold.
  • No alert was sent in the last Wait before repeating hours (default 12). This stops the same problem filling your inbox while you fix it.

You get one message when a cycle goes badly, never one per URL.

Only a real MISS counts as cold. Bypass, no cache and unknown answers do not. See Reading the cache answer.

What the email says

Section titled “What the email says”

The subject reads like “[Your site] Cache warmer: 52% of warmed pages came back cold”. The message gives the cold share against your threshold, the number of URLs measured, and how many were cold on desktop and on mobile, followed by the usual causes.

The dashboard shows the last alert under Rhythm > Last alert.

What to check

Section titled “What to check”

A healthy site sits in single digits. A high cold share usually has one of three causes. Check them in this order.

  1. A full pass takes longer than your cache lifetime. Pages expire before they are warmed again. Raise the batch size, shorten the worker interval, or turn on the quiet window. If the dashboard also says “A full pass takes more than a day”, this is the cause.

  2. Something in front of WordPress stopped caching. Check your cache plugin and the response headers by hand. A plugin that purges the cache on every save is a common culprit.

  3. The mobile user agent no longer matches your cache’s rules. The mobile requests then land in the desktop bucket, and the mobile copy stays cold. Check Mobile user agent under Settings, and compare the desktop and mobile columns under Recently warmed on the dashboard.

The status line on the dashboard

Section titled “The status line on the dashboard”

Even without email alerts, the line at the top of the dashboard warns you. It reads Pages are not staying warm when the cold share is above your threshold, and Warming has stopped when the worker has not run for too long. See The dashboard.