I’m a student in IT, in switzerland and I have to try to optimize the time request (XQuery) in a Tamino XML server.

I already do some test and I obtein some results that I don’t understand. (I join the request and the results).

I would like to know why the times request do not increase on the same way of the number of response and why the time request of the second request (11_rep) is always smaller than the the time request of the first one (2_rep). Have you an idea ?

Thanks in advance

R
Request.doc (64 KB)

Rejane,

without going into details (I could not anyway) two attempts to answer your questions:

- the requests do not increase with number of results: This is good Database behaviour. The time to find the results does not grow linear with number of results or number of documents in the database. Good indexing is the reason here.

- second request is faster: Caching. The Database caches index and data used in previous queries. Second querie uses cache (memory) instead of Hard disk. Much faster. To get usable results for your test use some repetitions for each query and use the average, even better to take out extreme results.

Hope this help,

Timm

(which tool are you using? MS WAS?

Thanks for your help.

the tool is JMeter from Jakarta. It’s very usefull to test different request (http, ftp, LDAP etc.) You can download it on this adress :

http://jakarta.apache.org/site/binindex.cgi#jmeter_binaries

R