Information Technology Reference
In-Depth Information
Catalog browse
12
Max response time = 10 sec
10
8
6
4
2
0
1
50
100
150
200
Number of round trip transactions
Figure 9.6
Round trip performance for peak catalog workload
9.5.3 Saturday Peak Workload Mix
Because all three transaction groups perform below (faster than) their required response
times at peak workload, it is appropriate to attempt a workload mix of the same volume
of transactions. Your fi rst thought might be to throw all three transaction groups into
the same performance measurement window and see if they continue to demonstrate
performance below their requirements. The problem with this approach is that if one
or more of the transaction groups' response times go above the requirement maximum,
you do not have many clues as to what in the mix changed the performance.
Consider a deliberate ramp-up of each transaction group into the mix in an order that
refl ects the business priority of the workload. For example, the business priority of the
Saturday workload is purchasing. So the suggested transaction group mix sequence for
the Saturday workload is (1) log on/log off, (2) purchase steps, and (3) catalog search.
Figure 9.7 shows the results of ramping up the log on/log off to peak workload,
then launching purchase steps to peak workload.
Log on
Purchase
8
8
Purch max
7 s
7
7
6
6
5
5
4
4
Log on max
3 s
3
3
2
2
1
1
1
50
100
150
200
250
300
350
400
450
500
Number of round trip transactions
Figure 9.7
Round trip performance for peak logon + purchase workload mix
Search WWH ::
Custom Search