[Vortex] Implementation question
"Robert M. Münch"
robert.muench at robertmuench.de
Sat Feb 12 10:04:41 CET 2011
Hi, I know that Vortex makes heavy use of threads to handle concurrency but I'm not sure about the internals. (And, yes I haven't scanned through the code to answer the following questions...)
Are there are any profiling data about the memory consumption pattern? I mean thread-based networking is relatively inefficient and very difficult to use. Systems using this in average allocate 2mb thread stacks for each connection. How is this done in Vortex?
Or does Vortex tell the operating system (through epoll, kqueue, /dev/poll, or select) that it should be notified when a new connection is made, and then it goes to sleep. If someone new connects, then it executes the callback and each connection only uses a small heap allocation.
Robert M. Münch
Mobile: +49 177 245 2802
More information about the Vortex