Made in USA: Enterprise Application Services

Ringing PhoneCall Today!817-210-4042

SAP Performance Management Solution

SAP Performance Management Solution

With Ayoka’s application development services, companies are able to monitor and improve the performance of their core business processes, including supply chain management systems, customer relationship management systems, and e-commerce systems used by B2B and B2C customers. Ayoka develops SAP performance management solutions for companies that rely on SAP.

For the SAP performance management solution below, performance refers to the ability to fulfill customer’s expectations for response time and throughput, such as viewing products or purchasing them from the shopping cart.

SAP Performance Management Solution – Business Problem:

A large manufacturer and distributor of consumer products internally developed e-commerce systems using SAP’s ABAP programming language. As online traffic increased, with more B2B and B2C customers making purchases through the SAP driven website, customers began to experience significant performance issues. Ayoka was contracted to design and implement an SAP performance management solution that would improve their websites’ performance and users’ experience.

This company was using multiple websites on the same back-end SAP server. SAP requires every web page on their site to be associated with a corresponding SAP transaction. This requires a log in to be initiated by the browser each time a user browses to a new page. Therefore, as the user navigates through the client website, each request is handled as a separate call to the SAP server. This setup caused a high stress on the back-end server during peak web traffic hours.

SAP Performance Management Solution – Technical Resolution:

In order to reduce the load on the SAP server, Ayoka developed an SAP performance management solution that used server-side caching. This allowed pages with static data elements and relatively static server responses to remain in the cache, alleviating the load issue. Through the SAP performance management solution, it was no longer necessary to send a request to SAP for static pages.

Load Testing Results:

To begin load testing for the SAP performance management solution and generate the maximum load on the client’s SAP server, Ayoka utilized JMeter to simulate the client-defined number of users. Table 1 below shows the actions and expected results of the simulated user while browsing through the client’s two websites (website 1 and website 2). The simulated user begins on the website home page, then proceeds to click on category and product links allowing for user think time (a designated period of time between user actions). The two websites are tested concurrently with 1100 virtual users on website 1 and 1000 user’s on website 2. Over a period of 2100 seconds, each of the 2100 virtual users are loaded one per second. The websites are tested with the SAP performance management solution (caching solution) and without (non-caching solution).

Table 1: Actions and Expected Results of Simulated Users (SAP Performance Management Solution)

Step Description Expected ResultUser Think
Time(s)
1Open website home pageHome page is displayed1
2Click on a category linkCategory page is displayed10
3Click on a product linkProduct details page is displayed10
4Click on a static pagePage is displayed10

The results from load testing both websites with and without the SAP performance management solution (caching solution) are displayed below in Tables 2 and 3. Overall, the Ayoka SAP performance management solution made the average page load time three times faster than the non-cached solution. In addition, the implementation of the SAP performance management solution resulted in a drop in the average error rate from 20.18% to 0.0%.

Table 2*: Baseline Load Testing Results for SAP Performance Management Solution (Non-Cached site)

Website 1 (1100 simulated users)

Sample LabelReport CountAverage (ms)Report MedianReport 90% LineReport MinReport MaxReport Error %Report RateReport Bandwith
Home Page110024222442838257612.181.0026.61
Click Category Link110045126077416250921.641.0063.85
Click Product Link11003202224559247401.001.0054.42
Click Static Page11003351427772647800.551.0033.48
TOTAL66003131727670509235.895.83175.64

Website 2 (1100 simulated users)

Sample LabelReport CountAverage (ms)Report MedianReport 90% LineReport MinReport MaxReport Error %Report RateReport Bandwith
Home Page100023821442638212312.301.0017.96
Click Category Link100054243776327739230.701.0080.20
Click Product Link100043338773619341770.401.0063.19
TOTAL30004044027233841774.472.94158.09

Table 3*: Load Testing Results for SAP Performance Management Solution (Cached site)

Website 1 (1100 simulated users)

Sample LabelReport CountAverage (ms)Report MedianReport 90% LineReport MinReport MaxReport Error %Report RateReport Bandwith
Home Page11001331281491104530.001.0089.57
Click Category Link11001221121441057200.001.0056.28
Click Product Link11001381311591106170.001.0048.99
Click Static Page1100111102129956810.001.0028.69
TOTAL6600114110141707200.005.83217.51

Website 2 (1100 simulated users)

Sample LabelReport CountAverage (ms)Report MedianReport 90% LineReport MinReport MaxReport Error %Report RateReport Bandwith
Home Page100013412815510613520.001.0074.77
Click Category Link10001371261631075070.001.0064.44
Click Product Link10001341241541096900.001.0047.95
TOTAL300013512615610613520.002.94183.44

SAP Performance Management Solution – ROI:

The SAP performance management solution allowed the e-commerce system to accommodate more web traffic during peak online shopping season. It decreased page load time and enhanced the user’s online experience. In addition, the SAP performance management solution enabled the manufacturer to reduce the number of customers switching from online transactions to telephone and reduce attrition, i.e. decrease the number of users abandoning sales due to unsatisfying performance.

* Below are definitions of the terms used in the SAP performance management solution results tables above.

TermDefinition
Report CountNumber of simulated users
Average (ms)The average page load time for a set of results
Report MedianThe time in the middle of a set of results
Report 90% Line90% of the samples took no more than this time
Report MinThe shortest time for the samples with the same label
Report MaxThe longest time for the samples with the same label
Report Error %Percent of request with errors
Report RateRequests per second
Report BandwidthThroughput measure in kilobyte per second