On 05/20/2016 05:16 PM, Dave Taht wrote:
We had found some pretty major performance problems on this hardware as of a few months ago. I am curious if they still exist? A whole bunch of benchmarks went by on the cerowrt-devel list, also.
Sorry I'm late on a response. I've been buried with work lately and haven't been paying much attention to mailing lists.
They were: 1) broken local ethernet network stack - running 4 copies of netperf against it - one would grab all the bandwidth and the other three barely even start or hang completely.
I confirmed your behavior on the 1900ACS using OpenWRT. I opted for flent rrul_be and rrul tests rather than multiple netperfs though.
I'm happy to rejigger my build to be LEDE, rebuild, install, and run the test again.
2) No BQL in the ethernet driver
Dunno about this one.
3) Wifi was horribly overbuffered in general 4) Wifi would crash against flent's rrul test.
I certainly saw these behaviors as well but I've not run rrul against the WiFi in quite a while and I seem to remember a wireless firmware update happening since I last ran the test.
I'll re-run and see. -- Richard A. Smith _______________________________________________ Lede-dev mailing list Lede-dev@lists.infradead.org http://lists.infradead.org/mailman/listinfo/lede-dev
