Learn how customers use SHA 9.23 to get more analytic, control, prediction…

I am very excited to share the latest and greatest in our new Service Health Analyzer (SHA) 9.23 with you. As I noted in my previous post, you can get SHA up and running with less than two minutes investment.

 

In this blog, I want to show you how you will be able to use several of the new features to gain more out of your SHA, more control, more prediction and more analytics.

 

If you aren't currently using Service Healh Analyzer, you can download it here.

 

Analytics capabilities 24/7

 

We have a new “Analytic CI” that allows you to gain analytics insights regardless of whether there’s an anomaly or not. Customers are using this tab to view the current status of their application, host or software element, see the metrics and gain insight into their metric behavior.

 

This tab is also widely used in POC as you can start seeing data within a few hours of your SHA installation, you will be able to view all the metrics that SHA monitors and begin to perform correlation analysis by using SHA advanced correlation tools or by using SHA graphs to overlay metrics for comparison.

 

More control over your SHA analytics

 

One of the world’s largest telco providers uses the new “Ignore Metric” feature to reduce their false alarms and to gain better focus in their SHA. This customer uses SHA for one of their large SAP applications, which has more than 5,000 users, because this application had a lot of abnormal metrics. By using the new feature in SHA, the customer was able to fine-tune SHA to better fit their application.

 

 Untitled.png

[Fig. 1: SHA Ignore Metric Options]

 

Another example of how SHA offers you more control over analytics is a unique case where one of the leading mobile telecommunications equipment and services provider uses SHA over their own business metrics. The customer requested to get alerted on metrics that drop more than 10%, even if this drop is momentary or within the baseline sleeve.

 

To accommodate this need, we use the new feature of SHA 9.23 that allows the customer to control the abnormality definition of each metric and enter their own logic by writing a simple groovy rule (Figure 3):

Groovy example.png

[Fig. 2: Groovy rule to control how you define a metric as abnormal]

This rule is now available in the product itself as an example.

 

Some metrics of HP Service Health needed to be analyzed with more contexts on certain days. Since we launched SHA, we have been approached by many customers asking us to support baseline for special days and holidays. For example,December 25 (Christmas Day) is obviously not a regular day and we should not expect it to behave as any other day. In our research we found out that:

1. To build a baseline for this date, we would need 3 years of data! Customers usually don’t have this data in their data warehouse, but even if they do, most applications will not behave the same across this long timeframe.

2. Although each application behaved differently on this special days, customers can often articulate what is the expected behavior and often this behavior can be derived from the regular behavior of the application. For example, the expected response time in Christmas is twice as high as a regular day, or the user expects the same response time on average but greater deviation (and willing to be more tolerant).

 

Using the groovy capability in SHA 9.23 we were able to address this need by writing a simple groovy rule (Figure 4):

Groovy example 2.png 

[Fig. 3: Groovy script example to accommodate special dates with abnormal behavior]

 

You can also use your SLA calendar to create and specify your own special days.

 

Predictive events for special sets of metrics

 

SHA was designed and built to support large IT organizations and analyze complex applications. As such, it applies several algorithms. One of them states that if in a given application only one metric is abnormal, then it is false alarm. However, there are several customers — among them one of largest online gambling businesses — that approached us to get predictive events for a special set of metrics, even if only one of this metric exceeded its baseline.

 

To support this use case we add a new setting that allows you to define these special metrics.

 

 setting.png

[Fig. 4: Infrastructure Settings ]

 

 

Trend Analysis

 

SHA now calculates and detects existing trends in your metrics. In the example below, we are able to establish using Diagnostic data that there is a memory leak in the system that is causing the heapTotal size to grow every day.

 trend.png

[Fig. 5: Trend analysis help you identified harmful trends in your IT]

 

To Summarized SHA 9.23 new features allow you to get more out of your SHA:

You can now use SHA capability 24/7 even at time there is n anomaly.

You have more control over your SHA to truly customized it to meet your special needs.

You have trend analysis that help you predict and visualized unwanted degradation in your services.

 

Get more information

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the Community Guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
Showing results for 
Search instead for 
Do you mean 
About the Author
Featured


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation.