Site24x7 polling logic during Monitor outage

Site24x7's intelligent polling logic during Monitor outage

For Internet service monitors, all paid plans, excluding the Enterprise plans allows you to monitor your websites from 8 test locations (1 primary and 7 secondary locations). For Enterprise plans, monitoring can be performed from up to 16 locations. On purchasing a custom plan, you get to monitor from all Site24x7 test locations. 

Site24x7 uses a predictable polling from a location logic to ensure a uniform and intelligible graph. On the contrary, other vendors use a round-robin methodology, which gives too many spikes between polls because of latency differences due to different geographical conditions. You can pick different sets of test locations for different monitors from our list of 130 geographical locations. Site24x7 allows monitoring from the primary location at a poll frequency, as low as one minute. Monitoring from secondary locations is performed every 30 minutes, by default. However, if you need secondary polling at a much lower frequency than this, you must contact our sales team.

Site24x7's location recheck logic during an outage

Our recheck logic is primarily based on the downtime rules set in your threshold profiles. In general, the primary location polls based on the check frequency set by you (can be as low as one minute); and the secondary locations poll, every 30 minutes (called Hourly Poll - Polling from all secondary Locations twice an hour), by default. In case the primary is DOWN, there will be immediate rechecks from 3 (by default) locations/based on the preset downtime rules in threshold configuration. This is a random selection of a location that we do from among the locations that are mapped in your location profile.

The rechecks from the secondary location will be done as per the set logic. Let's say monitor A has 1 primary and 7 secondary locations. Downtime rules configured are 3. If the primary is DOWN, there will be an immediate recheck from all the 7 secondary locations. If DOWN status is received from any two secondary locations, we will trigger an immediate alert without waiting for the response from the other 5 locations. However, polling from the remaining 5 locations will continue.

In case the recheck is set at 1, then we will alert immediately after the primary fails.

Location Rechecks will be automatically removed if the monitors are in DOWN status for more than 3 Hours. Hourly Poll will also be revoked if the monitors are in DOWN status for more than 10 days.

Uptime Steroids for Advanced Monitors

Uptime steroids are an added status check for advanced monitors like Web Transaction Monitor, Web Page Speed(Browser) and Web Transaction(Browser). This is to facilitate aggressive uptime monitoring, where status checks will be performed for the first URL in the sequence every 1 min. For Web Page Speed(Browser) monitor, the configured URL will be checked every minute.

If it identifies any change in status, it will trigger the monitor's actual data collection irrespective of the set polling frequency and identify the monitor's actual status. The actual monitor status data will be captured in the Monitor's Log Report. However, you'll not be able to view any performance metrics for these checks. Hence, the same will be highlighted as a (-) in the Log Report.
Note: If you wish to enable this option for your existing Web Transaction (Browser) Monitor, you can use the Edit option (you can access this by clicking on the hamburger icon in Monitor Summary Page) and then Save it. For disabling this option you can contact  Site24x7 Support.

Aggressive checks

When a monitor's status changes from UP to DOWN, aggressive rechecks will be performed at a certain frequency until the next poll, irrespective of the set polling interval. It will go into regular polling mode thereafter. Aggressive checks will be done only from the primary location. But if any status change is detected, it will immediately trigger an overall data collection from all the configured locations including the primary location. First recheck will be scheduled 30 seconds after detecting the first down time, irrespective of the poll interval configured for the monitor. Rechecks after that will depend on the poll frequency configured for the monitor, as shown below.

Check the table for the next recheck time based on  "down duration" and poll interval.
Note: Down duration stands for the time difference between current time and the last down time of the monitor. 

      Poll Interval
           Down Duration
  Next Poll Time
Any interval
< 1 minute

After 30 seconds

3 minutes
< 3 minutes
 
After 1 minute

5 minutes
 
< 3 minutes

< 5 minutes

After 1 minute

After 2 minutes
10 minutes

< 3 minutes

< 10 minutes

After 1 minute

After 3 minutes
15 minutes
< 5 minutes

< 15 minutes

After 2 minutes

After 3 minutes

>= 30 minutes

< 7 minutes


< 30 minutes

After 3 minutes


After 7 minutes

Note: If the down duration is more than 30 minutes or the configured poll interval, then the aggressive recheck will stop.
An alert will be triggered when there is a change in the status of the monitor from DOWN to UP.















    • Related Articles

    • Different types of polling for VMware resources

      Site24x7 supports monitoring for different VMware resources like vCenter, clusters, ESX/ESXi host, virtual machines (VMs), datastores, resource pools, and snapshots using On-Premise Poller. You can monitor the discovered VMware child resources using ...
    • Polling logic in Web Transaction (Browser) monitors

      Say you've set a check frequency of 5 minutes for your web transaction monitors and the first poll takes place at 10.00 AM. The last poll time in the monitor summary page will be updated when the entire data collection for that poll is completed. ...
    • What is Polling in Site24x7?

      Site24x7 captures the uptime and critical performance status of your websites, servers, and web applications. To achieve this, Site24x7 persistently checks the availability and collects relevant performance data from your monitored resources at ...
    • How are the average values calculated in Global Benchmark Report and Monitor Group Performance Report?

      Response time data in a Monitor Group Performance report and Monitor Group Global Benchmark Report is calculated using different logic. As a result, you may notice a slight deviation in the average values deciphered from these reports. Monitor Group ...
    • Polling frequency of monitors polled from On-Premise Poller

      The polling of monitors is done according to the configured poll frequency in the monitor. Data is pushed every five minutes from the On-Premise Poller. Poll now requests for monitors in the On-Premise Poller are pushed instantly after polling. Data ...