| || |
I am currently testing a datawarehousing product which is in it's initial stages of production.The problem is that it's a big system and i'm the only one testing it.I'm giving the details of what it's all about.Just let me know of the areas which you people think call for greater attention.
This is *******.Sorry for not posting this query earlier and now saying it's *******.Now the details.
All the machines within an IP range will be discovered;their hardware & OS detected,the applications,files and databases in these machines discovered according to what has been specified in the master database(for ex, if you give .txt as an entry,all text files will be searched for);metadata retriveal can be done on all the schemas discovered---this will give the list of tables along with their column attributes and constraints,keys,indexes;data retrieval can be done on the individual tables;any change in the metadata/data can be monitored through metadata monitoring/data monitoring;numeric analysis like sum,mean,variance etc, can be done on the numeric fields of all table data retrieved.
All this can be done only by administrators but there can be more than 1 admin asking for the same thing or different things at the same time.Further,the system is supposed to be OS independent and is supposed to discover data across a wide range of databases.
I think i've done a good job till now but frequent UI changes and incremental builds are now taking their toll on me.Would like to know the most significant areas and crash conditions you feel might arise here.Please feel free to clarify doubts you might have because i'll probably be following only this thread for some time now.Me,all alone in a data warehouse.
Re: Datawarehousing product
Are these major functions of the application:
If you take a risk-based approach towards the testing then what would be the riskiest problems? Perhaps retrieving data from machines that were out of the IP range? (This could be considered unauthorized access on someone else's machine.) Or would improper OS detection be risky? With a risk-based approach it depends on what the application is doing with the data and what the business values.
If you want to hit the major functions, then I see many test cases to cover connectivity, different client configurations, file formats and data formats.
Re: Datawarehousing product
Yes,all those except hardware discovery are major functionalities of the system.
Yes,the riskiest problems i have considered are machines out of the IP range,incorrect OS detection and discovering data from one machine while the system has been asked to do it from some other machine.Glad that only database discovery is happening erroneously and that too if there are more than one instances of the same databse installed in the network.
Lastly,most of the problems i have found are in the areas of connectivity,different configurations and different data formats.
After going through your reply,i assume i'm doing pretty well.
By the way,we just got the delivery date extended by a couple of weeks by the client.A big relief!!