Maintenance Period

The Maintenance Period feature allows you to designate a period of time, daily or weekly, for predefined hours for which you want to automatically stop alerting for some monitor(s). Typically, this will be during server maintenance or software upgrades.

Note: the Maintenance Period feature must be set before the downtime event to take effect. Otherwise, you will continue receiving alerts during maintenance.

If you set up a maintenance period, you will stop receiving problem alerts during the specified time interval.

However, receiving recovery alerts will not be stopped during the maintenance period. If the problem occurs before the set maintenance period you will receive a problem alert, and if your monitor recovers during the maintenance period, you will receive a recovery alert (see the examples below).

Your monitoring will not be stopped during the scheduled maintenance period. You will just stop receiving problem alerts and:

  • For all Server-Device, Application and custom monitors.
    No monitoring data will be included in your charts and tables if you choose so (by checking the Exclude failures from charts and tables box).
  • For other monitor types.
    NOKs will not be included in your charts and tables if you choose so (by checking the Exclude maintenance periods checks from charts and tables).

To set a maintenance period, click the icon in the top right corner of the monitor to open the context menu, and select Maintenance.

The Maintenance Periods window will open.
image001

Under Maintenance Type, select the option you want by clicking the respective radio button:

  • Select for this monitor if you want the Maintenance Period to apply to this monitor only.

– Alternatively, depending on the monitor type, select:

  • For uptime monitors of the group – the maintenance period will affect all Uptime monitors in the group this monitor belongs to.
  • For transaction monitors of the group – the maintenance period will affect all Transaction monitors in the group this monitor belongs to.
  • For full page load monitors of the group – the maintenance period will affect all Full Page Load monitors in the group this monitor belongs to.
  • Agent (for all associated monitors) – the maintenance period will affect all monitors that are using the same Monitis Agent.
  • For all monitors of the group – the maintenance period will affect all Server-Device, Application and Custom monitors which are included in the same group that this monitor belongs to.

image003

Select your time zone from the Timezone combo box.

Select the Period type:

  • Daily time schedule – to set the maintenance period to be applied daily within a specific time range.
  • Weekly – to set the maintenance period to be applied during a specific weekday range within a specific time range.
  • Custom – to set the maintenance period to be applied during a custom date/time period.

Set up your maintenance period by selecting the appropriate values from the combo boxes.

Check the Exclude failures from charts and tables checkbox if you want failures that will occur during the maintenance period not to be visible in charts and tables.

In case of Server-Device, Application and custom monitors the checkbox label will show Exclude maintenance periods checks from charts and tables, and if checked you will see no monitoring data for the maintenance period in charts and tables.

Click on the Add period button. The new maintenance period will be added in the list.
2016-03-28_1329

To delete a maintenance period, find the maintenance period in the list and click on the red cross icon next to it.

Click Done when finished.

NOK Counting During Maintenance

While a scheduled maintenance prevents problem alerts from being sent out, the system still “counts” NOK events during that time. Once the maintenance is over and if the assigned Alert Rule is still met, the system will send out an alert.

An example below will help you understand this better.

Example:

The User has an Alert Rule (AR) set on an uptime monitor M. The AR threshold is to alert if M fails from 2 locations simultaneously.

The User also sets a daily maintenance from 18:30 to 19:00.

At 18:45 M fails from location L1. Than at 18:59 M fails from location L2.

The AR threshold is crossed (failure from 2 locations simultaneously), but problem alert is not sent (because the maintenance is still on).

At 19:05 check from L1 fails again and the system sends out an alert (AR threshold is crossed, the maintenance is off).

It might seem that the system has fired an alert after failure from single location. However, this is the way the Maintenance feature works, and correct system behavior.

Downtime Calculation during Maintenance

Downtime, as shown in your recovery alerts and Alert History is calculated based on the assigned Alert Rule as time difference between the first problem alert (Alert Rule is met) and the recovery alert (Alert Rule is no longer met) (see Downtime Calculation).

Because of that, depending on whether the failure and recovery have occurred during the set maintenance period, the maintenance period time will count or not in downtime calculation.

The examples below will help you understand this better.

  1. If the first problem alert was sent before the maintenance period has started, and the problem has recovered before the maintenance period is over, the part of the maintenance period up to the problem recovery will be included in downtime calculation.

Example:

The user has an Alert Rule (AR) set on monitor M. The AR threshold is to alert if M fails from 2 locations simultaneously. The user also sets a daily maintenance from 18:30 to 19:00.

At 18:15 M fails from location L1. Then at 18:20 M fails from location L2. The AR threshold is crossed (failure from 2 locations simultaneously), and problem alert is sent to the user (because the maintenance is off).

At 18:35 M recovers from L1, and the system sends a recovery alert (maintenance is on, but unlike problem alerts recovery alerts are sent during maintenance) with the downtime of 15 min (from 18:20 to 18:35).

  1. If the first problem alert was sent before the maintenance period has started, and the problem has recovered after the maintenance period is over, the maintenance period will be included in downtime calculation.

Example:

The user has an Alert Rule (AR) set on monitor M. The AR threshold is to alert if M fails from 2 locations. The user also sets a daily maintenance from 18:30 to 19:00.

At 18:15 M fails from location L1. Then at 18:20 M fails from location L2. The AR threshold is crossed (failure from 2 locations simultaneously), and problem alert is sent to the user (because the maintenance is off).

At 19:05 M recovers from L1, and the system sends a recovery alert (maintenance is off) with the downtime of 45 min (from 18:20 to 19:05).

  1. If the problem occurs during the maintenance, problem alerts are not sent. When the maintenance is over and the system sends a problem alert the maintenance period will not be included in downtime calculation.

Example:

The user has an Alert Rule (AR) set on monitor M. The AR threshold is to alert if M fails from 2 locations.

The user also sets a daily maintenance from 18:30 to 19:00.

At 18:35 M fails from location L1. Then at 18:50 M fails from location L2. The AR threshold is crossed (failure from 2 locations simultaneously), but alert is not sent (because the maintenance is still on).

At 19:05 check from L1 fails again and the system sends out an alert (AR threshold is crossed, the maintenance is off). At 19:10 M recovers from L2 and the system sends a recovery alert with the downtime of 5 min (from 19:05 to 19:10).