Other Checks

Besides uptime checks, StatusCake monitors how fast pages load, whether certificates are healthy, and whether your scheduled jobs are running. They are managed in the same way as uptime checks, with the verbs list, describe, create, update and delete.

Pagespeed checks

A pagespeed check loads a page and records how long it took, how big it was and how many requests it made. Alerts are optional thresholds, and 0 turns one off.

Terminal

$ cakectl pagespeed create --name "Landing page" \
    --website-url https://www.example.com \
    --check-rate 3600 --region UK \
    --alert-slower 3000 --alert-bigger 2048

Flag	Meaning
--website-url	The URL of the page to test.
--check-rate	Seconds between tests: 60, 300, 600, 900, 1800, 3600 or 86400.
--region	Where to test from: AU, CA, DE, FR, IN, JP, NL, SG, UK, US or USW.
--alert-slower	Alert if the page takes longer than this many milliseconds.
--alert-bigger	Alert if the page is bigger than this many kilobytes.
--alert-smaller	Alert if the page makes fewer than this many requests.
--contact-groups	IDs of the contact groups to alert.

cakectl locations pagespeed lists the regions and what they are called.

As a file:

pagespeed.yaml

name: Landing page
website_url: https://www.example.com
check_rate: 3600
region: UK
alert_slower: 3000
alert_bigger: 2048
contact_groups:
- "1234"

SSL checks

An SSL check watches a certificate and alerts before it expires or when something is wrong with it.

Terminal

$ cakectl ssl create --website-url https://www.example.com \
    --check-rate 86400 --alert-at 7,14,30 \
    --alert-expiry --alert-broken --alert-mixed

Flag	Meaning
--website-url	The HTTPS URL whose certificate is checked.
--check-rate	Seconds between tests: 300, 600, 1800, 3600, 86400 or 2073600.
--alert-at	Days before expiry at which to alert.
--alert-expiry	Alert when the certificate expires.
--alert-broken	Alert when the certificate is broken.
--alert-mixed	Alert on mixed content, which is insecure content on a secure page.
--alert-reminder	Send reminders as expiry approaches.
--hostname	The hostname to present during the TLS handshake.
--follow-redirects	Follow redirects.

Heartbeat checks

A heartbeat check works the other way round. Rather than StatusCake testing your service, your job tells StatusCake that it ran. If a ping does not arrive within period seconds, the check is down. It suits work that has no address to test, such as a nightly backup or a queue consumer.

Terminal

$ cakectl heartbeat create --name "Nightly backup" --period 90000 \
    --contact-groups 1234 --tags production
$ cakectl heartbeat describe 123

describe shows the ping URL. Have the end of your job request it:

backup.sh

run-backup && curl -fsS --retry 3 "$HEARTBEAT_URL" > /dev/null

Because the ping URL is a secret, anyone who has it can report your job as healthy; list leaves it out, and only describe shows it. Give a job a period a little longer than its schedule, so that one slow run does not raise an alert.

Flag	Meaning
--period	Seconds to wait for a ping before alerting.
--host	The name of the hosting provider.
--tags	Tags.
--paused	Pause the check.

Finding and changing checks

Terminal

$ cakectl pagespeed list
$ cakectl ssl list
$ cakectl heartbeat list --status down
$ cakectl heartbeat update 123 --period 7200

Heartbeat checks can be filtered with --status, --tags and --match-any like uptime checks. As with uptime checks, listings leave out the uptime percentage unless you add --uptime. See Uptime Checks.

Next Steps

Decide who is told when a check fails, and when to stay quiet. See Alerting.