On 11/25/2012 06:49 AM, Alex Balashov wrote: > On 11/24/2012 05:41 PM, Sergey Okhapkin wrote: > >> Slow Java coupled with real-time SIP protocol? Hmm... Not viable to >> my point of view. > > Despite the stereotype, Java actually performs quite well, and is used > in highly concurrent, real-time environments like telephony > extensively. I think the stereotype about slow interpreted languages > belongs to the first era of Java (1996-2000(?)), back when it had > green threads and so on. > > Having said that, it is true that the JVM consumes a lot of resources > just to run at all, and that the bureaucratic fashions of Java are > chiefly an affectation of the enterprise world. However, once it > spools up and gets going, Java is quite fast, and adequate for SIP. > It has posted comparable benchmarks to C for many scientific computing > applications for years now. > > Initially, I was as surprised by this as you are. >
A problem with Java is garbage collection. Every now and then the application slows down. This is not good for real-time applications requiring millisecond precision. There are plenty of other problems with Java as well - startup is usually very slow. Applications are almost always bloated. And the Java programming culture of 'throw more iron at it' inevitably produces yet more bloat. _______________________________________________ SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
