I have a 12 CPU server with 3 4-port 82576 network cards (12 ports total using igb driver). It's running Debian 7 (3.2 kernel with the realtime patches).
I have a process on that system that is reading from a virtual tap interface every 20ms (using boost::asio) and writing udp packets out one of the igb ports. If I bring down or configure one of the other interfaces not related to that process, that process goes to sleep for a little over 200ms (fairly consistent, 205-220ms) and misses 10 of its timeslots. If I bring down the e1000 82574L interface on the motherboard, I don't see this problem. I tried running ftrace on my process, but it didn't tell me much. It seems that while I was bringing the interface down, the process switched CPUs, but there was a 200ms delay (closer to 300ms with ftrace running) between when it stopped on one CPU and when it started on another. I tried it with the stock 3.2 igb driver and also with the igb 5.1.2 driver, same problem. Any suggestions on how to troubleshoot? Thanks, Aaron ------------------------------------------------------------------------------ Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ E1000-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/e1000-devel To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
