<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic DPDK port issue in Embedded Intel® Core™ Processors</title>
    <link>https://community.intel.com/t5/Embedded-Intel-Core-Processors/DPDK-port-issue/m-p/197262#M481</link>
    <description>&lt;P&gt;Hello,&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;We are trying to run the &lt;B&gt;l3fwd &lt;/B&gt;DPDK sample with minor change; In some executions we are unable to read packets from the ports.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;l3fwd Sample Execution Details&lt;/B&gt;&lt;/P&gt;&lt;P&gt;The program reads packets using the &lt;B&gt;rte_eth_rx_burst&lt;/B&gt; (same as in the original sample), while the &lt;B&gt;send_burst&lt;/B&gt; frees mbufs structures without sending the packets. &lt;/P&gt;&lt;P&gt;It prints an rx counter aggregated by the returned &lt;B&gt;rte_eth_rx_burst &lt;/B&gt;value (same as the &lt;B&gt;l2fwd&lt;/B&gt; sample code does) and also prints the rx reads/drops statistics from the NIC.&lt;/P&gt;&lt;P&gt;The sample doesn't perform IP checksum nor does it use RSS.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;Problem Details&lt;/B&gt;&lt;/P&gt;&lt;P&gt;When we execute the sample using 11 connected ports (0xFF77) with traffic, while reading from 5 ports (0x171), we get the following results:&lt;/P&gt;&lt;P&gt;On some executions all the ports can be read (rte_eth_rx_burst is able to read packets), while on other executions some ports can't be read though they have traffic.&lt;/P&gt;&lt;P&gt;When the problem occurs all the ports within the same NIC won't read packets (although they all have packets sent to them).&lt;/P&gt;&lt;P&gt;Once one of the above behaviours occurred, it will be replicated continuously until the next machine reboot (init 6)&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;Our Environment Specifications:&lt;/B&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;System:&lt;/B&gt; x 3650 &lt;B&gt;M5&lt;/B&gt; with the following &lt;B&gt;X710&lt;/B&gt; interfaces: 3 &lt;B&gt;QUAD&lt;/B&gt; connected to PCI 0-3. 1 DUAL connected to PCI 5 (not being used)&lt;/P&gt;&lt;P&gt;&lt;B&gt;CPU:&lt;/B&gt; 4 cores are not isolated, the rest are. we are using isolated cores to run the sample&lt;/P&gt;&lt;P&gt;&lt;B&gt;MEM: &lt;/B&gt;we have 8 huge pages of 1G each.&lt;/P&gt;&lt;P&gt;&lt;B&gt;OS:&lt;/B&gt; REHL 6.5&lt;/P&gt;&lt;P&gt;&lt;B&gt;DRIVERS: &lt;/B&gt;1.3.46 for i40e&lt;/P&gt;&lt;P&gt;&lt;B&gt;DPDK:&lt;/B&gt; versions 1.8.0 / 2.1.0&lt;/P&gt;&lt;P&gt;&lt;B&gt;SAMPLE EXECUTION PARAMS:&lt;/B&gt; l3fwd 0c 0x3E -n 4 -- -p 0x171 -P --config="(0,0,1),(4,0,3),(5,0,4),(6,0,5),(8,0,2)"&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;please see the attached "modified l3fwd sample" (main.c).&lt;P&gt;&lt;/P&gt;&lt;P&gt;any help would be appreciated.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;thx,&lt;/P&gt;&lt;P&gt;Anderson.&lt;/P&gt;</description>
    <pubDate>Tue, 10 Nov 2015 23:26:32 GMT</pubDate>
    <dc:creator>sregi1</dc:creator>
    <dc:date>2015-11-10T23:26:32Z</dc:date>
    <item>
      <title>DPDK port issue</title>
      <link>https://community.intel.com/t5/Embedded-Intel-Core-Processors/DPDK-port-issue/m-p/197262#M481</link>
      <description>&lt;P&gt;Hello,&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;We are trying to run the &lt;B&gt;l3fwd &lt;/B&gt;DPDK sample with minor change; In some executions we are unable to read packets from the ports.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;l3fwd Sample Execution Details&lt;/B&gt;&lt;/P&gt;&lt;P&gt;The program reads packets using the &lt;B&gt;rte_eth_rx_burst&lt;/B&gt; (same as in the original sample), while the &lt;B&gt;send_burst&lt;/B&gt; frees mbufs structures without sending the packets. &lt;/P&gt;&lt;P&gt;It prints an rx counter aggregated by the returned &lt;B&gt;rte_eth_rx_burst &lt;/B&gt;value (same as the &lt;B&gt;l2fwd&lt;/B&gt; sample code does) and also prints the rx reads/drops statistics from the NIC.&lt;/P&gt;&lt;P&gt;The sample doesn't perform IP checksum nor does it use RSS.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;Problem Details&lt;/B&gt;&lt;/P&gt;&lt;P&gt;When we execute the sample using 11 connected ports (0xFF77) with traffic, while reading from 5 ports (0x171), we get the following results:&lt;/P&gt;&lt;P&gt;On some executions all the ports can be read (rte_eth_rx_burst is able to read packets), while on other executions some ports can't be read though they have traffic.&lt;/P&gt;&lt;P&gt;When the problem occurs all the ports within the same NIC won't read packets (although they all have packets sent to them).&lt;/P&gt;&lt;P&gt;Once one of the above behaviours occurred, it will be replicated continuously until the next machine reboot (init 6)&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;Our Environment Specifications:&lt;/B&gt;&lt;/P&gt;&lt;P&gt;&lt;B&gt;System:&lt;/B&gt; x 3650 &lt;B&gt;M5&lt;/B&gt; with the following &lt;B&gt;X710&lt;/B&gt; interfaces: 3 &lt;B&gt;QUAD&lt;/B&gt; connected to PCI 0-3. 1 DUAL connected to PCI 5 (not being used)&lt;/P&gt;&lt;P&gt;&lt;B&gt;CPU:&lt;/B&gt; 4 cores are not isolated, the rest are. we are using isolated cores to run the sample&lt;/P&gt;&lt;P&gt;&lt;B&gt;MEM: &lt;/B&gt;we have 8 huge pages of 1G each.&lt;/P&gt;&lt;P&gt;&lt;B&gt;OS:&lt;/B&gt; REHL 6.5&lt;/P&gt;&lt;P&gt;&lt;B&gt;DRIVERS: &lt;/B&gt;1.3.46 for i40e&lt;/P&gt;&lt;P&gt;&lt;B&gt;DPDK:&lt;/B&gt; versions 1.8.0 / 2.1.0&lt;/P&gt;&lt;P&gt;&lt;B&gt;SAMPLE EXECUTION PARAMS:&lt;/B&gt; l3fwd 0c 0x3E -n 4 -- -p 0x171 -P --config="(0,0,1),(4,0,3),(5,0,4),(6,0,5),(8,0,2)"&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;please see the attached "modified l3fwd sample" (main.c).&lt;P&gt;&lt;/P&gt;&lt;P&gt;any help would be appreciated.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;thx,&lt;/P&gt;&lt;P&gt;Anderson.&lt;/P&gt;</description>
      <pubDate>Tue, 10 Nov 2015 23:26:32 GMT</pubDate>
      <guid>https://community.intel.com/t5/Embedded-Intel-Core-Processors/DPDK-port-issue/m-p/197262#M481</guid>
      <dc:creator>sregi1</dc:creator>
      <dc:date>2015-11-10T23:26:32Z</dc:date>
    </item>
    <item>
      <title>Re: DPDK port issue</title>
      <link>https://community.intel.com/t5/Embedded-Intel-Core-Processors/DPDK-port-issue/m-p/197263#M482</link>
      <description>&lt;P&gt;Hello Anderson,&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Thank you for contacting the Intel Embedded Community.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;We suggest you verify that you have followed the guidelines stated in chapter &lt;A href="http://dpdk.org/doc/guides/sample_app_ug/l3_forward.html"&gt;http://dpdk.org/doc/guides/sample_app_ug/l3_forward.html&lt;/A&gt; 13. L3 Forwarding Sample Application of the DPDK 2.1.0 documentation.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Please let us know if this information is useful to you.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Best Regards,&lt;/P&gt;&lt;P&gt;Carlos_A.&lt;/P&gt;</description>
      <pubDate>Wed, 11 Nov 2015 17:31:01 GMT</pubDate>
      <guid>https://community.intel.com/t5/Embedded-Intel-Core-Processors/DPDK-port-issue/m-p/197263#M482</guid>
      <dc:creator>CarlosAM_INTEL</dc:creator>
      <dc:date>2015-11-11T17:31:01Z</dc:date>
    </item>
  </channel>
</rss>

