AlertSite provides several monitoring modes. A monitoring mode controls the location usage order, whether the locations coordinate error detection with each other, and which locations send alerts.
You can select the desired mode in your monitor settings.
Monitoring Mode | Primary | Global Notify | Global Verify | Round Robin | SLA (MultiPOP) (used for SLA monitoring) |
---|---|---|---|---|---|
Locations | 1+ | 1+ | 1+ | 2+ | 2+ |
Checking from multiple locations | In parallel | In parallel | In parallel | One after another | In parallel |
Rotating through location pool | – | Yes | Yes | Yes | Yes |
Local retry on error | Yes, if the monitor has local retry enabled. | ||||
Error check from another location | Yes | – | Yes | – | – |
When alerts are sent | Primary location finds an error | Any location finds an error | Any location finds an error and the error is verified from another location. | 2 or more locations find an error within the same run | Errors are found at 1, 2, 3, or all locations within the same run. The number of locations is specified by the SLA Alerts option in your AlertSite preferences. |
Private locations supported | Yes | Yes | Yes | – | – |
Notes:
-
Legacy plans (Performance Pro, SLA) allow rotated locations only for the Round Robin and SLA (MultiPOP) modes.
-
Round Robin monitors are compatible only with public locations.
-
In the Primary mode, errors detected at non-primary locations are included in the reports, but alerts are not sent.
-
On the Usage-Based Monitoring plan, for monitors that use both public and private locations:
-
Location rotation is enabled by default for public locations.
-
Private locations run every interval.
-
The primary location must be a public one.
-
-
In the Global Verify mode, if a monitor uses both public and private locations, monitoring runs only on public locations.
Error Verification
Before sending an error alert, all monitors repeat the check to avoid false positives. Depending on the options, a monitor re-checks from the same location and another location.
The error verification procedure is as follows:
-
If the monitor’s Enable Local Retry option is selected, the monitor checks again from the same location.
-
If the second check does not confirm the error, the monitor reports the warning state. A check from another location is not performed.
-
If the second check confirms the error, and either:
-
the monitoring mode is Global Verify
-
the monitoring mode is Primary, and the error occurred at the primary location
-
the error is NOT a full page object error (status codes 7121, 7122, or 7123)
In these cases, the monitor also checks from an alternate remote location.
-
-
If the remote location also detects an error, an alert is triggered.
Considerations for Usage-Based Monitoring
On the Usage-Based Monitoring plan, the default mode for new monitors is Global Verify, set to rotate locations, with 1 location per interval. Changing the monitoring mode, run interval, the number of locations used, or rotation options affects the credit usage rate.
For example, if you have a monitor with 4 locations:
-
The Primary mode will use 4 credits on every check – one for each location.
-
The Global Verify mode with Rotate Locations enabled and 1 locations per interval will use 1 credit per check, because it will rotate through the 4 locations one at a time for each test.
-
The Global Verify mode with Rotate Locations disabled will use 4 credits per run interval, because it will check from all 4 locations at the same time.
How many locations you can check your website from. The maximum number of locations depends on your monitoring plan.
On each run, the monitoring locations check your website either in parallel or one after another.
Rotation means the monitor uses a subset of its location pool (say, 2 out of 10 locations) on every run, cycling through the locations. This way you can monitor from many locations, while limiting the number of checks on each run.
If rotation is not used, the monitor checks from all of its locations every time.
You enable or disable rotation in the monitor options.
When a monitor finds an error from some location, it repeats the check from the same location to avoid false positives.
When a monitor finds an error from some location, it repeats the check from another location to avoid false positives.
Alerts can be triggered on errors at any location, a specific location or at multiple locations at the same time.
A private node is a private monitoring station installed on your local network, allowing you to monitor internal applications that are not accessible from the public Internet.
Some monitoring modes are not compatible with private nodes.