On 2009-04-17, Devesh Sharma <devesh28@xxxxxxxxx> wrote: > Hi michi, > Its is using a TCP/IP protocol, program is Intel's MPI benchmark IMB, > and CPU is surely not a bottelneck because its Quad core Quad socket > machine with 64GB of physical mem. its a proprietary device for > Systems Area Networks used for cluster computing. But the main problem > that I have observed is, my device do not get any packets to post > during this data size thats why I am trying to look beyond my device > related functions. You should be able to see from your test tool whether the packets drop occurs in the sent end or receive end. Put the TCP latency aside. Just test with the layer 2 packets to see if there's any packet loss. If no, there's something to do with TCP. > > > On Fri, Apr 17, 2009 at 1:49 PM, Michael Blizek ><michi1@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote: >> Hi! >> >> On 11:03 Fri 17 Apr , Devesh Sharma wrote: >>> Hello list, >>> >>> I have written a network device driver, and to validate it, I am >>> running a bandwidth measurement tool on it, >>> I am encountering a strange drop in bandwidth when data size reaches >>> to 4096, bandwidth figures drop form 24 Mbps (for 2k data size) >>> to .43 Mbps. what can be the problem, the MTU of my device is 2044. Is >>> it some how related to functioning of hard_start_xmit? >> >> What program/protocol do you use for testing + what kind of device is >> this? The CPU is not a bottleneck, is it? >> >> -Michi >> -- >> programing a layer 3+4 network protocol for mesh networks >> see http://michaelblizek.twilightparadox.com >> >> > > -- > To unsubscribe from this list: send an email with > "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx > Please read the FAQ at http://kernelnewbies.org/FAQ > > -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
