On Wed, Aug 22, 2012 at 9:17 AM, girish kumar <[email protected]> wrote:
> 1. We have a scenario where there are multiple publishers(threads) and > multiple subscribers(threads). We are not able to understand a model in zero > MQ, that works for this scenario. Bind the side that has the least less dynamic IP addresses, connect the others to them. If you really have a cloud of pubs and subs all dynamically coming and going, use a device in the middle (forwarder) with a known IP address. > - As zero MQ sockets are not thread safe, can we make zero MQ sockets > threadsafe by applying explicit locks?. It's a bad idea from several aspects, leads to poor design and unstable code. 0MQ sockets are not thread safe because well-designed 0MQ apps use messages to communicate, not locks. > 2. Another problem is we are running some performance tests for best > results. We would like to know the highest possible numbers(latency and > throughput) achieved with zero MQ. You can use the performance tests in the perf/ subdirectory. -Pieter _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
