topic DL380Gen9 Server PCI NUMA allcation in ProLiant Servers (ML,DL,SL) https://community.hpe.com/t5/proliant-servers-ml-dl-sl/dl380gen9-server-pci-numa-allcation/m-p/6992257#M160307 <P>Hi,</P><P>&nbsp;</P><P>We have a DL380Gen9 Server with 24CPU's and two threads of 12CPU each, loaded with Ubuntu[14.04.1]. We have loaded the PCI slot with NIC's and when checked for NUMA node for PCI's slots, i could see all the NIC's are mapped to single NUMA node (/sys/class/net/*/device/numa_node). Which should be expected, am i correct?</P><P>Is there any way we can allocate NUMA node for PCI from diffrent Node?</P><P>&nbsp;</P><P># numactl --hardware<BR />available: 2 nodes (0-1)<BR />node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 24 25 26 27 28 29 30 31 32 33 34 35<BR />node 0 size: 128870 MB<BR />node 0 free: 107714 MB<BR />node 1 cpus: 12 13 14 15 16 17 18 19 20 21 22 23 36 37 38 39 40 41 42 43 44 45 46 47<BR />node 1 size: 129021 MB<BR />node 1 free: 101244 MB<BR />node distances:<BR />node 0 1<BR />0: 10 21<BR />1: 21 10<BR /># numactl --show<BR />policy: default<BR />preferred node: current<BR />physcpubind: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47<BR />cpubind: 0 1<BR />nodebind: 0 1<BR />membind: 0 1<BR />#</P><P># cat /sys/class/net/*/device/numa_node<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />#</P> Wed, 27 Dec 2017 08:43:01 GMT Vishruth 2017-12-27T08:43:01Z DL380Gen9 Server PCI NUMA allcation https://community.hpe.com/t5/proliant-servers-ml-dl-sl/dl380gen9-server-pci-numa-allcation/m-p/6992257#M160307 <P>Hi,</P><P>&nbsp;</P><P>We have a DL380Gen9 Server with 24CPU's and two threads of 12CPU each, loaded with Ubuntu[14.04.1]. We have loaded the PCI slot with NIC's and when checked for NUMA node for PCI's slots, i could see all the NIC's are mapped to single NUMA node (/sys/class/net/*/device/numa_node). Which should be expected, am i correct?</P><P>Is there any way we can allocate NUMA node for PCI from diffrent Node?</P><P>&nbsp;</P><P># numactl --hardware<BR />available: 2 nodes (0-1)<BR />node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 24 25 26 27 28 29 30 31 32 33 34 35<BR />node 0 size: 128870 MB<BR />node 0 free: 107714 MB<BR />node 1 cpus: 12 13 14 15 16 17 18 19 20 21 22 23 36 37 38 39 40 41 42 43 44 45 46 47<BR />node 1 size: 129021 MB<BR />node 1 free: 101244 MB<BR />node distances:<BR />node 0 1<BR />0: 10 21<BR />1: 21 10<BR /># numactl --show<BR />policy: default<BR />preferred node: current<BR />physcpubind: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47<BR />cpubind: 0 1<BR />nodebind: 0 1<BR />membind: 0 1<BR />#</P><P># cat /sys/class/net/*/device/numa_node<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />0<BR />#</P> Wed, 27 Dec 2017 08:43:01 GMT https://community.hpe.com/t5/proliant-servers-ml-dl-sl/dl380gen9-server-pci-numa-allcation/m-p/6992257#M160307 Vishruth 2017-12-27T08:43:01Z Re: DL380Gen9 Server PCI NUMA allcation https://community.hpe.com/t5/proliant-servers-ml-dl-sl/dl380gen9-server-pci-numa-allcation/m-p/7000418#M161834 <P>Hello&nbsp;<a href="https://community.hpe.com/t5/user/viewprofilepage/user-id/1892063">@Vishruth</a>!</P><P>Sorry for delay with response. Is your question still valid?</P><P>Thank you!</P> Thu, 22 Mar 2018 10:58:14 GMT https://community.hpe.com/t5/proliant-servers-ml-dl-sl/dl380gen9-server-pci-numa-allcation/m-p/7000418#M161834 denis_b 2018-03-22T10:58:14Z