Reference Guide

Alerts roll-up

Alerts roll-up is the process of consolidating the alerts based on the circumstances that are possessed at the time of generation

of platform alerts and at the time of moving the live alerts to historical alerts. As of now, the roll-up is applicable only for the

platform alerts.

The roll-ups are at two different places:

Live alerts roll-up for platform alerts

The roll-up of alerts happens if there are multiple alerts that are generated for the same condition ID within a span of 30 sec

interval. In 30 seconds window, the initial five events generate as many alerts, and the later events are rolled up and generate a

single alert depicting the number of times the event occurred.

Closing of rolled-up alerts (automatically through system)-If the clear rolled-up alerts generate for these rolled-up alerts,

then these rolled-up alerts can be closed. If system generated clear events are received more than five within the 30 seconds

window, then clear roll-up alert is generated.

Alerts roll-up if the notifications service was down

If the notification service is down momentarily, and some events are generated gradually during that time, then on the service

restart, it processes all the events, and the same roll-up logic is applied as the events are consumed through the service from

broker within 30 seconds.

Historical alerts

Historical alerts provide a view to the user to look into the occurrences of the particular alerts in the past. Historical alerts

contain the consolidated data of the generated alerts over a period of last six months. The alerts are aggregated and stored in

the historical table. The user has been provided with default filters for the 7 Days, 14 Days, and 30 Days. These filters include

the data for the number of days in the filters along with the current day data. For example, Last 7 Days data contains the data

for last 7 days and the current days.

Customized date filter

The custom dates filter is provided where user can get data for a customized interval. That customized date filter contains the

data from 00:00:00 hours start date to 23:59:59 hours end date. For example, if the start date is selected as 1 Jul 2020 and end

date as 15 Jul 2020, then user sees the alerts from Jul 1 00:00:00 hours to Jul 15 23:59:59 hours.

Aggregation for alerts

While adding to the historical alerts, the platform alerts with the same Condition ID, Resource and State are aggregated as one

with the addition of a column Count which represents the number of times that particular alert is generated in the time window

filter which the user has applied.

The hardware alerts are aggregated based on the Message ID field with the addition of a column Count which represents the

number of times that particular alert is generated. The logs field in the properties panel contains the creation date and last

updated date for these occurrences. The rolled-up live alerts are stored as in historical alerts table as these alerts are already

rolled-up.

A single row represents an alert with its particular state (OPEN or CLOSED), number of occurrences on a given day, and

whether it is enabled or disabled.

So, based on this scenario, a particular condition_id can have the following entries for the historical alerts:

● Enabled-OPEN rolled-up historical alert

● Enabled-CLOSED rolled-up historical alert

● Disabled-OPEN rolled-up historical alert

Events and Alerts