Tune monitoring sensitivity

UptimeHawk checks your websites and servers from monitoring stations in North America, Europe, Asia/Pacific and elsewhere. A single failed check is not treated as an outage: when one location cannot reach your site, other locations check again to confirm it before an outage is recorded and alerts are sent. This filters out problems at the monitoring station or on one network path.

The monitoring profile decides how many locations confirm an outage, how long each check waits for an answer and how long to wait between confirmations. It is a trade-off:

• More sensitive: short outages, such as a reboot, are caught and reported quickly, but you may get the occasional false alarm.
• Fewer false alarms: an outage must be confirmed from more places, so very short outages may not be reported and DOWN alerts arrive slightly later.

The default suits most websites. Change it only if you get false alarms, or if short outages you care about go unreported.

Choose a profile

uptimehawk.com/en/monitoring-profile.php

123

Close

123

1. Open Monitoring and select Monitoring Profile.

2. Choose a profile:

   Profile	Locations	Timeout	Delay	Best for
   System default marker 1	Currently the same as preset #2			Most websites
   Preset profile #1 (high sensitivity)	2 random	30 sec	0 sec	Catching every short outage, with occasional false alarms
   Preset profile #2 marker 2 (medium sensitivity)	2 random	30 sec	20 sec	General use
   Preset profile #3 marker 3 (low sensitivity)	4 random	60 sec	20 sec	Sites that get false alarms on the default
   Custom	You choose	You choose	You choose	Checking from specific locations

3. Select Update.

This sets the profile for your whole account. To use a different profile for one website or server, choose Monitoring Profile from its action list on Monitoring Overview.

Build a custom profile

Choose Custom to pick the settings yourself:

• Monitoring location #1 to #4: where checks run, in order. Choose a specific city, a random location in a region (Random North America, Random Europe) or Random All. If the first location sees a failure, the next confirms it after the delay, and so on; an outage is only recorded if every location used confirms it. Pick at least two different locations.
• Timeout: how long each check waits for an answer, from 15 to 180 seconds.
• Delay between checks: how long to wait before the next location confirms a failure, from 0 to 60 seconds. A longer delay skips very short outages.
• Backup monitoring location: used when a location you chose is unavailable.

Monitoring from locations near your visitors gives figures that match what they see. Quick Test shows how your site responds from every location; see Troubleshoot with a quick test.

Change the HTTP headers sent with checks

Website and HTTP checks send standard browser headers. If your firewall or security plugin blocks the checks, or your site serves different content by language or browser, adjust them:

1. Open Monitoring and select HTTP Headers.

2. Change a header’s value, or select +Add More to add one. In values, #USERNAME# is replaced with your UptimeHawk username and #MONITORING_HOSTNAME# with the monitoring host’s name.

3. Select Update. Restore Default puts the original headers back.

These headers apply to every website and server. To change them for one website or test only, use HTTP headers in its advanced options.

Related

• Set up downtime alerts (to only be alerted after several failed checks)
• Troubleshoot with a quick test