From: Glenn H. <thr...@ma...> - 2005-04-12 16:26:00
|
Off hand, I don't know of any performance testing that has been done. The rest of my answers are embedded. On 4/11/05 10:58 AM, "ar...@be..." wrote: > Dear all, > > we are in the phase of decision making, whether we shall switch from > Bugzilla to Mantis or not and need your help. > Specially we want to know, whether mantis is stable enough for 20 > concurrent users with a request delay of 10 seconds (each user will start > 10 seconds after the previous user had begun her request) . I'm currently > testing this scenario on suse linux which is hosted on a vmware virtual > mashine with JMeter. My VM has 512 MB of RAM and a 1,8 Mhz CPU. The > response time is more than 3 seconds, if I repeat the test 100 times. If I > repeat the tests only 10 times, I will be shown with a average response > time of 0,5 seconds, which is accptable. > > Now my questions: > > 1) has anybody performed such a testing scenario and could confirm these > results? > 2) where is the limit accepting user requests? This is probably in the number of concurrent threads that the web server can support (as well as memory). > 3) how many thousands / ten thousands of records can Mantis manage with > MySQL? We have had reports of up to about 10,000 records in a database. The mantisbt tracker has about 5000 issues and 2000 users. > 4) Is it possible to use Oracle? Support for Oracle is just being developed. Testers and feedback are welcome. There are some items in the forum and tracker on this. > 5) Is Mantis tunable, if yes how? What aspects do you want to tune? > 6) Who is Using Mantis for more than 20 concurrent users? > 7) Can I import my Bugzilla reports into Mantis? There isn't an off the shelf solution, but I'm sure that someone can develop one for you. > > Thanks in advance for possible answers. > > Best Regards, > Arash Kaffamanesh |