Comments on: MSI-X – the right way to spread interrupt load http://www.alexonlinux.com/msi-x-the-right-way-to-spread-interrupt-load Fri, 30 Jul 2010 05:08:02 +0000 hourly 1 http://wordpress.org/?v=abc By: Alexander Sandler http://www.alexonlinux.com/msi-x-the-right-way-to-spread-interrupt-load/comment-page-1#comment-23988 Alexander Sandler Sun, 04 Jul 2010 20:38:11 +0000 http://www.alexonlinux.com/?p=1632#comment-23988 <a href='#comment-23986' rel="nofollow">@ninez</a> Thanks for sharing your experience and for a warm comment. Please come again! :-)<div class="comment-remix-meta"><a href="#" class="replyto" onclick="replyto('23988','Alexander Sandler'); return false;">Reply</a> - <a href="#" class="quote" onclick="quote('23988','Alexander Sandler','<a href=\'#comment-23986\' rel=\"nofollow\">@ninez<\/a>\r\nThanks for sharing your experience and for a warm comment. Please come again! :-)'); return false;">Quote</a></div> @ninez
Thanks for sharing your experience and for a warm comment. Please come again! :-)
Reply – Quote
]]> By: ninez http://www.alexonlinux.com/msi-x-the-right-way-to-spread-interrupt-load/comment-page-1#comment-23986 ninez Sat, 03 Jul 2010 16:23:21 +0000 http://www.alexonlinux.com/?p=1632#comment-23986 MSI-X is great, and is also now used by defualt in the linux kernel. 2.6.33 and onwards! no more sharing irq on my laptop. great article!<div class="comment-remix-meta"><a href="#" class="replyto" onclick="replyto('23986','ninez'); return false;">Reply</a> - <a href="#" class="quote" onclick="quote('23986','ninez','MSI-X is great, and is also now used by defualt in the linux kernel.\r\n\r\n2.6.33 and onwards!\r\n\r\nno more sharing irq on my laptop.\r\n\r\ngreat article!'); return false;">Quote</a></div> MSI-X is great, and is also now used by defualt in the linux kernel.

2.6.33 and onwards!

no more sharing irq on my laptop.

great article!

Reply – Quote
]]> By: telenne barz http://www.alexonlinux.com/msi-x-the-right-way-to-spread-interrupt-load/comment-page-1#comment-23908 telenne barz Mon, 17 May 2010 10:06:47 +0000 http://www.alexonlinux.com/?p=1632#comment-23908 Hi Alex ! Once again, here's a nice article... Thanks for sharing your knowledges. For outbound packets, the Kernel builds a hash based on IP addresses and port numbers (source & destination, I suppose ?) in order to bind the corresponding flow to a given TX queue. I was wondering if the hash is build in the same manner for inbound packets / RX queues ? What I understand is that the driver is in charge of binding a given ingress flow to a given RX queue. Does that mean that the sysadmin cannot configure it a posteriori (with ethtool for instance) ? "Using interrupt affinity binding techniques you can bind certain interrupt vector to certain core" : can you please give us further details on how to to setup that ? Does that mean that each queue will appear as a particular device under /proc/interrupts ? Finally, did you hear about the TNAPI and PF_RING patches of Lucas Deri (http://www.ntop.org/TNAPI.html) ? If the MSI-X feature is already implemented in the concerned drivers (Intel igb, igbx), I don't catch what is the benefit of the TNAPI patch. What is your opinion about this ? Telenn<div class="comment-remix-meta"><a href="#" class="replyto" onclick="replyto('23908','telenne barz'); return false;">Reply</a> - <a href="#" class="quote" onclick="quote('23908','telenne barz','Hi Alex !\r\n\r\nOnce again, here\'s a nice article... Thanks for sharing your knowledges.\r\n\r\nFor outbound packets, the Kernel builds a hash based on IP addresses and port numbers (source & destination, I suppose ?) in order to bind the corresponding flow to a given TX queue. I was wondering if the hash is build in the same manner for inbound packets \/ RX queues ?\r\n\r\nWhat I understand is that the driver is in charge of binding a given ingress flow to a given RX queue. Does that mean that the sysadmin cannot configure it a posteriori (with ethtool for instance) ?\r\n\r\n\"Using interrupt affinity binding techniques you can bind certain interrupt vector to certain core\" : can you please give us further details on how to to setup that ? Does that mean that each queue will appear as a particular device under \/proc\/interrupts ?\r\n\r\nFinally, did you hear about the TNAPI and PF_RING patches of Lucas Deri (http:\/\/www.ntop.org\/TNAPI.html) ? If the MSI-X feature is already implemented in the concerned drivers (Intel igb, igbx), I don\'t catch what is the benefit of the TNAPI patch. What is your opinion about this ?\r\n\r\nTelenn'); return false;">Quote</a></div> Hi Alex !

Once again, here’s a nice article… Thanks for sharing your knowledges.

For outbound packets, the Kernel builds a hash based on IP addresses and port numbers (source & destination, I suppose ?) in order to bind the corresponding flow to a given TX queue. I was wondering if the hash is build in the same manner for inbound packets / RX queues ?

What I understand is that the driver is in charge of binding a given ingress flow to a given RX queue. Does that mean that the sysadmin cannot configure it a posteriori (with ethtool for instance) ?

“Using interrupt affinity binding techniques you can bind certain interrupt vector to certain core” : can you please give us further details on how to to setup that ? Does that mean that each queue will appear as a particular device under /proc/interrupts ?

Finally, did you hear about the TNAPI and PF_RING patches of Lucas Deri (http://www.ntop.org/TNAPI.html) ? If the MSI-X feature is already implemented in the concerned drivers (Intel igb, igbx), I don’t catch what is the benefit of the TNAPI patch. What is your opinion about this ?

Telenn

Reply – Quote
]]>