Last modified on February 19, 2018

AlertSite dashboards and reports display measurement data at varying times, depending on server and network availability. The table below displays latencies from the time at which an event occurs until the time the data appears in the respective screen:

Feature Expected Latency Worst Case
Dashboard < 3 minutes 5 minutes
Console < 3 minutes 5 minutes
Reports (except usage reports) < 10 minutes 20 minutes
Usage reports < 3 minutes 5 minutes
Report API < 10 minutes 20 minutes

Notes:

  • Assuming there is sufficient connectivity to pull the data; data is pulled immediately after an error.

  • Latencies could be longer if there is no connectivity or if extensive maintenance is occurring.

  • Timing starts when the monitoring server has finished monitoring for the test, not when starting the test.

  • Verification and second attempt add additional latency.

Data Processing Frequency

  • Console and Dashboard:

    • Measurement data is pulled from every monitoring host in parallel every two minutes by two database servers in two different geographical locations.

    • The data is imported and processed directly upon arrival.

  • Customer Database Servers

    • Data is pulled in parallel every two minutes from all monitoring servers, with a pause to ensure the entire set is finished.

    • A separate process runs every 5 minutes to insert the pulled data.

Issues that May Cause Delays

  • Nightly Backups (DB Servers)

    • Backups are started each night at 04:45 GMT.

    • Additional CPU or disk load may increase delay.

  • Weekly Database Reorganization

    • Data beyond customer's data retention period is purged beginning at 12:00 GMT on Saturdays.

    • Database is re-optimized after purge for faster access.

  • Connectivity Problems

    • Data updates cannot occur if the database servers cannot connect to a monitoring server.

    • If connection occurs and bandwidth is poor, data updates may be delayed for up to 45 minutes (the worst case).

    • Connectivity issues are minimized by having two database servers at two different geographical locations.

  • Maintenance

    • Periodic maintenance of servers is a required process.

    • Maintenance may involve moving a large database to another server for better system performance.

See Also

AlertSite UXM Dashboard
About Reporting