[ovs-discuss] unable to map backing store for hugepages: Cannot allocate memory

Charlie Li charlie.li at gmail.com
Wed Aug 3 15:20:42 UTC 2016


Thanks Mark,

That is a very good point.

Just found out I forgot to add "--socket-mem 1024,0" after increasing the
page number from 1 to 4.

With "sudo ./ovs-vswitchd --dpdk -c 0x1 -n 4 --socket-mem 1024,0 --
unix:$DB_SOCK --pidfile --detach", the problem is gone.

Now I got:

qemu-system-x86_64: -netdev
type=vhost-user,id=net1,chardev=char1,vhostforce: chardev "char1" went up
qemu-system-x86_64: -netdev
type=vhost-user,id=net2,chardev=char2,vhostforce: chardev "char2" went up

Then the terminal is frozen.

How can I get into the console of the VM?

Regards,
Charlie

On Wed, Aug 3, 2016 at 9:24 AM, Kavanagh, Mark B <mark.b.kavanagh at intel.com>
wrote:

> >
> >Thanks Sugesh for your response.
> >
> >I have 4 1G hugepages allocated and the VM is requesting 3G memory.
> >
> >$ cat /proc/cmdline
> >BOOT_IMAGE=/vmlinuz-4.5.7-200.fc23.x86_64
> root=/dev/mapper/fedora--desktop-root ro
> >rd.lvm.lv=fedora-desktop/root rd.lvm.lv=fedora-desktop/swap
> default_hugepagesz=1G
> >hugepagesz=1G hugepages=4 rhgb quiet
> >
> >$ grep Huge /proc/meminfo
> >AnonHugePages:         0 kB
> >HugePages_Total:       4
> >HugePages_Free:        0
> >HugePages_Rsvd:        0
> >HugePages_Surp:        0
> >Hugepagesize:    1048576 kB
>
> Hi Charlie,
>
> Even though you have 4 pages mounted, none are available (see
> Hugepages_Free = 0, above).
>
> Have you tried unmounting and remounting the hugepages?
>
> Thanks,
> Mark
>
> >
> >Regards,
> >
> >Charlie
> >
> >On Tue, Aug 2, 2016 at 5:23 PM, Chandran, Sugesh <
> sugesh.chandran at intel.com> wrote:
> >
> >
> >Regards
> >_Sugesh
> >
> >From: discuss [mailto:discuss-bounces at openvswitch.org] On Behalf Of
> Charlie Li
> >Sent: Tuesday, August 2, 2016 4:28 PM
> >To: discuss at openvswitch.org
> >Subject: [ovs-discuss] unable to map backing store for hugepages: Cannot
> allocate memory
> >
> >Hi All,
> >
> >I am trying to use dpdkvhostuser to pass traffic to VM.
> >
> >Here is my basic system configuration.
> >
> >Host and VM OS: Fedora server 23
> >DPDK 2.2.0
> >OVS 2.50
> >QEMU 2.4.1
> >
> >When I tried to start the VM, it got the following error:
> >
> >qemu-system-x86_64: -object memory-backend-file,id=mem,size=3072M,mem-
> >path=/dev/hugepages,share=on: unable to map backing store for hugepages:
> Cannot allocate
> >memory
> >
> >I must did some thing wrong.
> >Any help is appreciated.
> >
> >Thanks,
> >
> >Charlie
> >
> >More details
>
> >---------------------------------------------------------------------------------------------
> >----------------------
> >
> >[cli at cli-desktop ~]$ grep Huge /proc/meminfo
> >AnonHugePages:         0 kB
> >HugePages_Total:       4
> >HugePages_Free:        0
> >HugePages_Rsvd:        0
> >HugePages_Surp:        0
> >Hugepagesize:    1048576 kB
> >
> >[cli at cli-desktop utilities]$ sudo ./ovs-vsctl show
> >4fcf27a6-edbd-4770-ab39-f440e3532bcc
> >    Bridge "br0"
> >        Port "vhost0"
> >            Interface "vhost0"
> >                type: dpdkvhostuser
> >        Port "vhost1"
> >            Interface "vhost1"
> >                type: dpdkvhostuser
> >        Port "br0"
> >            Interface "br0"
> >                type: internal
> >        Port "dpdk1"
> >            Interface "dpdk1"
> >                type: dpdk
> >        Port "dpdk0"
> >            Interface "dpdk0"
> >                type: dpdk
> >
> >[cli at cli-desktop utilities]$ sudo ./ovs-vsctl list-ports br0
> >dpdk0
> >dpdk1
> >vhost0
> >vhost1
> >
> >[cli at cli-desktop utilities]$ sudo ./ovs-ofctl dump-flows br0
> >NXST_FLOW reply (xid=0x4):
> > cookie=0x0, duration=61391.667s, table=0, n_packets=0, n_bytes=0,
> idle_age=61391, in_port=1
> >actions=output:3
> > cookie=0x0, duration=61391.650s, table=0, n_packets=0, n_bytes=0,
> idle_age=61391, in_port=2
> >actions=output:4
> > cookie=0x0, duration=61391.632s, table=0, n_packets=0, n_bytes=0,
> idle_age=61391, in_port=3
> >actions=output:1
> > cookie=0x0, duration=61391.616s, table=0, n_packets=0, n_bytes=0,
> idle_age=61391, in_port=4
> >actions=output:2
> >
> >[cli at cli-desktop utilities]$ sudo qemu-system-x86_64 -m 3072 -cpu host
> -hda
> >/home/cli/VM1/FC23.qcow2 -boot c -enable-kvm -pidfile
> /home/cli/VM1/vm1.pid -monitor
> >unix:/home/cli/VM1/vm1monitor,server,nowait -name 'FC23-VM1' -net none
> -no-reboot -object
> >memory-backend-file,id=mem,size=3072M,mem-path=/dev/hugepages,share=on
> -numa node,memdev=mem
> >-mem-prealloc -chardev
> socket,id=char1,path=/usr/local/var/run/openvswitch/vhost0 -netdev
> >type=vhost-user,id=net1,chardev=char1,vhostforce -device virtio-net-
>
> >pci,netdev=net1,mac=00:00:00:00:00:01,csum=off,gso=off,guest_tso4=off,guest_tso6=off,guest_ec
> >n=off,mrg_rxbuf=off -chardev
> socket,id=char2,path=/usr/local/var/run/openvswitch/vhost1 -
> >netdev type=vhost-user,id=net2,chardev=char2,vhostforce -device
> virtio-net-
>
> >pci,netdev=net2,mac=00:00:00:00:00:02,csum=off,gso=off,guest_tso4=off,guest_tso6=off,guest_ec
> >n=off,mrg_rxbuf=off --nographic -vnc :14
> >
> >qemu-system-x86_64: -object memory-backend-file,id=mem,size=3072M,mem-
> >path=/dev/hugepages,share=on: unable to map backing store for hugepages:
> Cannot allocate
> >m/emory
> >[Sugesh] May be you are out of hugepages? How much memory you are
> allocating for OVS-DPDK?
> >
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openvswitch.org/pipermail/ovs-discuss/attachments/20160803/11662013/attachment-0002.html>


More information about the discuss mailing list