UST SmartOps™ uses cognitive automation to intelligently and continuously accelerate business opportunities by reducing costs, streamlining efficient operations, and uncovering unrealised growth revenues for your business.

AIOps Release 2.5

 Release Note 

 Release Date: May 10, 2022 

Comparison between AIOps similarity and Hybrid correlation methods

Event driven email notifications

Introducing maintenance window and cooling period

Device status based on failure and recovery alerts

Display time zone as per user preference

Customized ticket thresholds using rules

Dashboard widgets for auto resolution analytics

Overview

AIOps 2. 5 release introduces multiple new features such as event driven email notifications, maintenance window and cooling period for devices, custom ticket thresholds and other enhancements in ticket management and auto-resolution.

Compare Correlation Methods

In the previous release, we introduced AIOps Similarity Correlation. In this release, we introduce a comparison between the two available correlation methods, to help decide which is the better choice for a customer. Comparison statistics are based on the latest 100 alerts and the current settings.

 

Comparison Metrics Description
Number of alerts used: Total number of alerts considered for comparison
Matched alerts: Total number of alerts that went into an existing cluster
Number of clusters: Total number of clusters formed
Cluster with maximum number: Total number of alerts present in the cluster which has maximum number of alerts
Cluster with minimum number: Total number of alerts present in the cluster which has minimum number of alerts
Number of rules/attributes used: Total number of rules used for correlation in case of hybrid correlation and number of attributes in case of similarity correlation
Mean: Arithmetic average of alert count in the cluster formed
Median: Arithmetic median of alert count in the cluster formed

 

Event driven email notifications

Introducing a generic framework for event-based notification within AIOps. Users will now be able to send email notifications during events such as ticket is created, cluster is acknowledged by user, etc.

Events for which email notification can be configured

  1. When ticket created for alert cluster
  2. When ticket created from acknowledgement alert cluster
  3. When ticket created manually
  4. When an alert cluster is acknowledged by the user
  5. When surge ticket is created

 

Introducing Maintenance Window & Cooling Period

Maintenance mode value ('ON' & 'OFF') will not be explicitly set by the user anymore. From this release, the user can define maintenance window for scheduled downtimes. After the maintenance window ends, a cooling period can be scheduled to define how long the device will require to be back to normal operation.

Cooling window can be set up at the project level and is an optional setting.

Maintenance window can be scheduled in bulk using the device inventory upload. For individual devices, the maintenance schedule can be updated via device details screen.

With this change, the API for updating maintenance mode ‘/updateDeviceMaintenanceStatus’ should not be used and will soon be decommissioned.

Device status based on failure and recovery alerts

The failure policy determines if an alert is a failure alert or not. The recovery policy will determine if an alert is a recovery alert or not. Device status will be determined by the absence/presence of failure alerts and matching recovery alerts for a device. In this context, matching refers to an exact match between the node name and alert metric of the failure and recovery alert.

Device status becomes UP when

  • All cluster(s) with failure alert(s) for a device is closed via auto closure or user actions or reverse sync (as existing).
  • For every open failure alert on the device, a recovery alert with the same alert metric is received with alert created time >= alert created time of the failure alert.
  • When there is a single matching recovery alert corresponding to multiple open failure alerts

 

Device status becomes DOWN when

  • On receiving a new failure alert, the device status will be updated to DOWN only when there is no matching recovery alert available with alert created time > = alert created time of the new failure alert

 

 

Sample Scenarios Device Status
1 FAIL alert is in unclosed cluster: no matching REC alerts DOWN
1 FAIL alert is in unclosed cluster: at least 1 matching REC alert in any cluster UP
2 FAIL alerts in unclosed cluster: 1 matching REC alert in any cluster (matches both FAIL alerts) UP
2 FAIL alerts in unclosed cluster: 2 matching REC alerts in any cluster (matches FAIL alerts separately) UP
2 FAIL alerts in unclosed cluster: 2 REC alerts in any cluster (matches only 1 FAIL alert) DOWN

 

Set up display time zone based on user preference

Users will now be able to set up their individual preferred time zone and format to be used for time fields displayed on screen. This will be useful for remote ITOps teams working in a different time zone than their customers.

You'll see this setting in effect on all time fields in all these areas:

  • Alert/Ticket listing
  • Configuration screens
  • Automation
  • Alert Download

 

The only exception will be the Grafana based dashboards, which will continue to show the local browser time.

Customized Ticket Threshold using rules

Ticket Threshold configuration is now customizable using policies and rules. This will allow admins to set different thresholds based on alert attributes. For example, admin can set up different thresholds for different types of alerts such as server alert, warning alert originating from same/different source.

Remove restriction on permitting only allowed fields in Ticket Templates

From this release onwards users can use any fields except the not allowed ones in the ticket templates.

Reverse sync of re-opened tickets from ITSM

From this release, when ticket gets re-opened in ITSM, it will be synced back to AIOps. AIOps will change the ticket status to Open and make it available for correlation.

Dashboard widgets for auto resolution analytics

Auto resolution analytics dashboard will now be available in all AIOps Desk & Resolve projects. This can be accessed from Manage Dashboards option.

  • (7181) Investec CORR error - Error in adding notification
  • (7239) LoanDepot : Correlation Error - "Failure in add_cluster_size"
  • (7488) NinetyOne - Alert missing
  • (7938) Muthoot: Error when downloading alert (empty data)
  • (7986) Muthoot: Saved filter showing 'No data available'
  • (8070) Muthoot- Advance filter is not working with resource name

Release Artifacts

Deployment Artifacts

For support and more enquiries, you can write a mail to us on feedback.smartops@ust-global.com