This is a web based application the flow of the application is as follows:
First we are using mercury load runner 8.1.
1). User will login.
2). Select appropriate work type.
3). Gives the necessary test inputs.
4). after getting the desired page update the necessary field and click apply. (Once you click on apply it will update the record and leads to work type selection page. Like this we need to do 30 times before logout)
5). Then logout.
We have now couple of hundred users and around 1200 test subjects. Each user has to do the update part of the application (Num.3 and 4) 30 times and log out. If any two, or more than two users want to update the same test subject (customer record), application doesnít allow. Users need to select unique test subjects. But we donít have more that 1200 subjects and we donít want to reduce 30 times update for each user.
In the load runner tool we have parameterization options like sequential, random and unique. We cannot select first two as we cannot update same record with multiple users. If we select unique we need to have more than 1200 test subject which we donít have.
In this situation please suggest me how to proceed further.
1. What version of LoadRunner (LR) or PerformanceCenter (PC) are you using? (specify which tool LR or PC)
2. What is the protocol you are recording?
3. If HTML - are you using HTML-Advanced with URLs or,
4. If URL mode:
- Concurrent groups, or
- Without Concurrent groups?
5. Which LoadRunner/PerformanceCenter feature (FPs) or service packs are you using?
(List details here)
6. VuGen Recording - are you using Old or New Recording Engine?
7. What kind of LoadRunner license do you have -
8. If subscription, is subscription current or expired?
(substitute your answer here)
9. What platform(s) (PCs) and Operating Systems (Windows-XP, etc.)? Include version and service packs (SP1 or 2, etc.)
WINDOWS XP, SERVER 2000
10. If you have filed a service request with HP/Mercury, what have they told you at this point with respect to your issue?
THEY HAVE ASKED TO SELECT NEXT ROW AS RANDOM AND UPDATE VALUE EACH OCCURANCE. I CAN DO THAT BUT THERE IS A POSIBILITY 2 OR MORE USERS CAN ACCESS SAME CUSTOMER RECORED AT THE SAME TIME WHICH APPLICATION DOESNíT ALLOW.
Does the scope of the test allow users to keep updating a single test subject, or limit himself to a block of six? and carry out 30 operations on those six subjects?
If yes, you have the options to do that in the parameter configuration window when you select Unique (define a block size and cycle thru the values and restart from begining when out of values)
If no, then the client either has to provide you the additional subjects or you have to alter your test scope.
[ QUOTE ]
In the load runner tool we have parameterization options like sequential, random and unique. We cannot select first two as we cannot update same record with multiple users
[/ QUOTE ]
Your concept here seems to be flawed and it depends a lot on the other factors like how hour you are running, how many iteration, whats the response time of the script etc.
For your question, fetch more data and its up to the data generation team to give you the required data for your testing.
I'd say run the test and see how many update errors you get. You'll probably get more than you'd like, but perhaps not enough to justify spending time on perfecting your test.
If you need a more perfect solution, I'd suggest you look at the Virtual Table Server option. This is an in memory database which can track cross-vuser data for you. You could place the record number of a record currently being accessed in the VTS and check for this every time a user iterates. This will, in essence, "lock" the record.
While dealing with VTS isn't rocket science, it will take some time for you to understand the first time. I'd say run your test, see if the error rates are acceptable, and if not, lobby for more time and look into VTS.
To follow up on what Dan said...
Data configuration is one of the major challenges of any load test. You must analyze what you are testing, and how many users you have to do this testing. Many times you must create locked in processes so that you can guarantee there are no conflicts between users.
For you problem, see if you can segregate your vusers in such a way that you can preclude you contention issue.