Weekend Testing 27 -- Test and Experience Report

Mission: To generate test ideas to test Google Buzz with quality criteria Performance.

Date: 06th March 2010, 3:05 PM IST.

To Deliver: Report of performance testing ideas for Google Buzz.

Tester: Ravisuriya

Context: Tester has been asked to look for test ideas to test performance quality criteria of Google Buzz. It was an opportunity and first time, a tester was working with performance quality criteria. Google Buzz was live and used by users for buzzing.

Started of with mission statement. I was not very sure what 'performance' meant in mission. With a question to WT session facilitator, assumed 'performance' here as "how quick and responsive it is". Further continued to collect details regarding the context in which 'performance' was considered here. But it remained as open question and was left to tester choice in session. With mental modeling myself as a tester who is looking out performance test ideas for Google Buzz, began to brainstorm.

Assumptions made while brainstorming:

• Was not sure what the ‘performance’ means in the test mission.
• A thought, does the performance mean – “how well was Google Buzz was performing among with similar other available social networking services”?
• Or how well the desired purpose of application satisfies the desired claims with given constraints by processing user actions and its throughput interval.
• Did assume for this session, the ‘performance’ is how quick and responsive it is.
• Various users are using Google Buzz at present, with no similar internet connectivity speed.

Few questions that I got while brainstorming about users and Google Buzz performance:

• Performance test ideas in what context?
• Who are the users and what kind of users are using Google Buzz?
• How often they have used and using Google Buzz? What are their observations and perceptions of performance?
• What devices did they use to browse the Google Buzz? Did this device too have any influence on their perception of Google Buzz performance?
• How long did they use Google Buzz in a stretch i.e., without any break?
• What environment did they have or do they have while using Google Buzz? Did it have any influence?
• Any tools are used to know or understand the performance parameters of application under test?
• How the performance is being measured? What are the units of measurement of performance in this context?

Brainstorming for Test Ideas:

• Identify the kind of users, their ages, business, purpose, environment, consistency of using Google Buzz. Few to mention here. This keeps growing as and when brainstorming and testing sessions is on.
• Identify the possible and potential contexts with each of this environment above said.
• Usage of analytics and log files of Google Buzz. The log files and analytical information regarding performance of application under test to have time of request processing and of throughput.
• Test on various possible hardware and *software configurations and network topologies. Note: *software and hardware may be that are installed on device from where Google Buzz is browsed and on Google Buzz server.
• Does Google Buzz can be accessed by mobile phones? If yes how the configuration and applications can have influence on time factor with Google Buzz responsiveness?
• Type of database Google Buzz makes use of.
• Handling the incorrect or invalid entries. How quick the application is responsive to user here?
• How quick does it handle if I keep buzzing with a very **small interval of gap between two buzz? Note: **small – will be decided on application’s minimum and maximum tolerance value.
• Does any other application using internet or network will have influence on Google Buzz’s performance?
• Study of features available or provided by Google Buzz. This might give more ideas how performances tests can be devised.
• Tests for knowing how the performance of application goes, when number of parallel users increased and decreased, number of transactions in a given time period varied to threshold and least, time consumed to recover from these two edges of variations etc.
• Performance of application when it's potential reaches to the extreme. What time it takes to handle the requests from clients?
• Type of server being used. Type of connections. Any data stored by application on client is not stored or stored partially – how the performance of application goes here.
• How Google applications being used by user simultaneously along with Google Buzz choke up the time. Need to identify such application(s) of Google and need to be tested along with Google Buzz.
• Other vast applications used over internet and Google Buzz.
• Maximum requests that can be handled at a given time. (This includes various actions using various features of Google Buzz).
• Usage of Google Buzz over ***network that is not comfort for using of application. Note: ***network – properties of network.
• If any tools or program used to look into the performance rating of application, what time is taken by tool to work on this? Does the time taken here by tool or program instructions have any influence?
• How the request of client is sent to server?

The test ideas I got during this session were of generic to performance. I will use these test ideas to develop specific performance tests. Interaction with Dr. Meeta Prakash showed how these ideas can be turned to be more specific to performance test ideas. Discussion session transcript is here. PDF document of my session report is here.