SCALABILITY TEST AND ANALYSIS

• US 20050216235A1
• Filed: 03/29/2004
• Published: 09/29/2005
• Est. Priority Date: 03/29/2004
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Hagiu, Costin, Butt, Hammad
• Granted Patent: US 7,010,465 B2
• Time in Patent Office: Days
• US Class Current: 702/186
• CPC Class Codes: G06F 11/3414 Workload generation, e.g. scripts, playback
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 11/3457 Performance evaluation by simulation
• CPC Class Codes: G06F 11/3476 Data logging G06F11/14, G06F11/2205 take precedence
• CPC Class Codes: G06F 2201/88 Monitoring involving counting

Abstract

Apparatuses and methods to test whether a multi-user system will provide satisfactory performance are described. Response times are logged for each individual user, and the measurements are aggregated together in a single file at the end of the test. For each action type, a graph is built that correlates the distribution of the response times as a function of the user load. A break point is determined for each action type at which a response time exceeds a predetermined threshold. By analyzing the different break points, the number of users that can be supported by the multi-user computer system is determined. Additionally, an optimal amount of memory may be determined to support a user load. The amount of memory required per user is computed based on the user load at the projected point where a line that is determined from page output peaks intersects the page input line.

Claims

1. A method for determining hardware configuration scalability of a multi-user computer system, comprising:

(A) loading a set of simulated actions grouped into action types and associated with a remote session;

(B) measuring and logging response times for the set of simulated actions;

(C) repeating (A) and (B) to obtain a plurality of response times;

(D) aggregating the plurality of response times to obtain aggregated results;

(E) splitting the aggregated results according to each action type; and

(F) determining an associated break point for each said action type, wherein the associated break point corresponds to a minimally acceptable degree of performance for each said action type.

3. The method of claim 1, further comprising:

(G) fitting a performance curve for an action type for a plurality of performance instances, wherein each performance instance corresponds to a response time in response to an occurrence of a simulated user action; and

(H) determining the associated break point from the performance curve.

10. The method of claim 1, further comprising:

(G) configuring a client terminal to reflect a set of user actions that are associated with a type of user.

11. The method of claim 1, further comprising:

(G) configuring the multi-user computer system to execute at least one application program that is utilized by a type of user.

12. The method of claim 11, wherein the type of user is selected from the group consisting of a knowledge worker and a data entry worker.

13. The method of claim 1, wherein the multi-user computer system comprises a terminal server.

14. The method of claim 1, wherein an initial response time of an action type is greater than a predetermined initial response time threshold and wherein the associated break point corresponds to an average response time having a predetermined degree of degradation.

15. The method of claim 1, wherein an initial response time of an action type is less than a predetermined initial response time threshold and wherein the associated break point corresponds to an average of the predetermined initial response time threshold and an average response time having a predetermined degree of degradation.

16. A computer-readable medium having computer-executable instructions for performing the method as recited in claim 1.

18. The method of claim 3, wherein (G) comprises:

(i) fitting the performance curve with a moving average for a predetermined number of performance instances.

2. The method of claim 2, further comprising:

(G) determining a number of users that the multi-user computer system can support from a plurality of associated break points.

4. The method of claim 2, further comprising:

(H) if the number of users does not equal or exceed a desired number of users, enhancing the computer system'"'"'s configuration; and

(I) repeating (A)-(G).

5. The method of claim 2, wherein the number of users corresponds to a processed break point, wherein the processed break point is an approximate average of all break points determined in (F).

6. The method of claim 2, further comprising:

(H) rank ordering the plurality of associated break points, wherein a first break point corresponds to a best performance and a last break point corresponds to a worst performance, wherein (G) comprises;

(i) selecting an n^th break point in response to (H), wherein the n^th break point corresponds to the number of users.

7. The method of claim 2, wherein a selected break point is an approximate weighted average of all break points determined in (F) and wherein the selected break point corresponds to the number of users.

8. The method of claim 2, wherein a processed break point corresponds to a minimum of all break points determined in (F) and wherein the processed break point corresponds to the number of users.

9. The method of claim 2, wherein a selected break point corresponds to a maximum of all break points determined in (F) and wherein the selected break point corresponds to the number of users.

17. A computer-readable medium having computer-executable instructions for performing the method as recited in claim 2.

19-31. -31. (canceled)

32. A method for determining hardware configuration scalability of a multi-user computer system, comprising:

(A) configuring a plurality of client terminals to reflect a set of user actions that are associated with at least one type of user;

(B) loading a set of simulated actions grouped into action types and associated with a remote session;

(C) measuring and logging response times for the set of simulated actions;

(D) repeating (B) and (C) to obtain a plurality of response times;

(E) aggregating the plurality of response times to obtain aggregated results;

(F) splitting the aggregated results according to each action type;

(G) determining an associated break point for each said action type, wherein the associated break point corresponds to a minimally acceptable degree of performance for each said action type; and

(H) determining a selected break point that corresponds to a minimum average response time, wherein the selected break point corresponds to a number of users that the multi-user computer can support.