>-----Original Message----- >From: Jeff Campbell [mailto:[email protected]] >Sent: Tuesday, September 07, 2010 10:25 AM >To: Ronciak, John >Cc: [email protected] >Subject: Re: [E1000-devel] 82574L - Multicast transmit >failing, causes performance issues > >John, > >> >> Thank you for the response. We did some further testing last night: >> >> -We tried compiling the driver without MSI/MSI-X support, >and that appeared >> to make no difference >> >> -Also, accessing the console via ssh is extremely sluggish >when a stream is >> being sent, it can take 3-4 seconds to respond to a CTRL-C >for example (or >> even CR), when dual 5 Mbps multicast streams are being output >> >> -The mouse is a bit sluggish on the console, not awful, but >noticeably >> impacted >> >> -No cpu load reported in userspace by top >> >> -We will shut down X and retest to eliminate that and >determine performance >> at the cli >> >> Stats will be forthcoming as well. >> >> -Jeff >> > >I should also clarify how we access the device. We have >physical console >access. Remote access for management is via eth0 on 192.168.1.0/24. > >All of the multicast traffic (in/out) is on eth1 on 10.239.0.0/28. > >eth0 is connected to VLAN 10 on the switch (admin/internet) > >eth1 is connected to VLAN 30 on the switch (video only) > >The switch is an HP Procurve 1800-24G. It is NOT IGMP aware, >and thus the >multicast is broadcast to all ports on VLAN30. Flow control >is enabled on >the switch ports in VLAN30. > >There are no streams, other than test streams, active during >the tests. The >switch is running the latest firmware and handles up to 40+ Mbps of >aggregate multicast streams without any problems. We >regularly have dual SD >(~5Mbps) and dual HD (~19 Mbps) streams running concurrently >without any >impairments. > >Is there any value in trying an earlier kernel version or is >that unlikely >the source of the problem? > >Is there any value in trying to reproduce the problem using multicast >streams under Windows (we would have to install a new HD and >install windows >to do this)? > >-Jeff > Hello Jeff,
I missed that John has been helping you with this, so ignore the stats requests I made. However, I think the lspci -vvvxxx output of both systems is still worth looking at, if you can provide it. Thanks, Carolyn Carolyn Wyborny Linux Development LAN Access Division Intel Corporation ------------------------------------------------------------------------------ This SF.net Dev2Dev email is sponsored by: Show off your parallel programming skills. Enter the Intel(R) Threading Challenge 2010. http://p.sf.net/sfu/intel-thread-sfd _______________________________________________ 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
