HI All,
After configuring hugepages of size 2048KiB at the host,I am not seeing any
improvement in the performance.Still after pumping more than 4Mpps i am
seeing this counter getting incremented
rx_no_dma_resources: 93356 at the host.
output of following commands are mentioned below-:
vpp# show interface
Name Idx State Counter
Count
VirtualFunctionEthernet0/a/0 1 up rx packets
177953727
rx bytes
11389034442
tx packets
177952680
tx bytes
16371646028
drops
1052
ip4
177952663
local0 0 down
vpp# show run
Thread 0 vpp_main (lcore 0)
Time 393.8, average vectors/node 1.00, last 128 main loops 0.00 per node
0.00
vector rates in 0.0000e0, out 1.0158e-2, drop 2.5394e-3, punt 0.0000e0
Name State Calls Vectors
Suspends Clocks Vectors/Call
VirtualFunctionEthernet0/a/0-o active 4 4
0 2.69e3 1.00
VirtualFunctionEthernet0/a/0-t active 4 4
0 4.72e3 1.00
acl-plugin-fa-cleaner-process event wait 0 0
1 1.64e4 0.00
admin-up-down-process event wait 0 0
1 7.39e3 0.00
api-rx-from-ring any wait 0 0
29 2.64e5 0.00
bfd-process event wait 0 0
1 6.42e3 0.00
cdp-process any wait 0 0
41 1.28e5 0.00
dhcp-client-process any wait 0 0
4 5.04e3 0.00
dpdk-ipsec-process done 1 0
0 2.89e5 0.00
dpdk-process any wait 0 0
132 6.89e4 0.00
error-drop active 1 1
0 2.35e3 1.00
fib-walk any wait 0 0
197 8.02e3 0.00
flow-report-process any wait 0 0
1 7.02e2 0.00
flowprobe-timer-process any wait 0 0
1 4.88e3 0.00
ikev2-manager-process any wait 0 0
394 4.81e3 0.00
ioam-export-process any wait 0 0
1 1.18e3 0.00
ip4-glean active 1 1
0 1.56e4 1.00
ip4-lookup active 4 4
0 6.53e3 1.00
ip4-rewrite active 3 3
0 3.02e3 1.00
ip6-icmp-neighbor-discovery-ev any wait 0 0
394 3.78e3 0.00
l2fib-mac-age-scanner-process event wait 0 0
1 1.22e3 0.00
lisp-retry-service any wait 0 0
197 8.37e3 0.00
lldp-process event wait 0 0
1 4.15e6 0.00
memif-process event wait 0 0
1 1.74e4 0.00
nat64-expire-walk done 1 0
0 1.01e4 0.00
send-garp-na-process event wait 0 0
1 1.29e3 0.00
snat-det-expire-walk done 1 0
0 1.05e3 0.00
startup-config-process done 1 0
1 5.97e3 0.00
udp-ping-process any wait 0 0
1 1.05e4 0.00
unix-cli-sockaddr family 1 active 0 0
298 7.72e6 0.00
unix-epoll-input polling 655279 0
0 1.56e6 0.00
vhost-user-process any wait 0 0
1 1.33e3 0.00
vhost-user-send-interrupt-proc any wait 0 0
1 1.06e3 0.00
vpe-link-state-process event wait 0 0
2 5.78e3 0.00
vpe-oam-process any wait 0 0
193 5.68e3 0.00
vpe-route-resolver-process any wait 0 0
4 7.33e3 0.00
vxlan-gpe-ioam-export-process any wait 0 0
1 2.31e3 0.00
---------------
Thread 1 vpp_wk_0 (lcore 1)
Time 393.8, average vectors/node 17.54, last 128 main loops 0.00 per node
0.00
vector rates in 4.5190e5, out 4.5190e5, drop 3.3698e0, punt 0.0000e0
Name State Calls Vectors
Suspends Clocks Vectors/Call
VirtualFunctionEthernet0/a/0-o active 10146115 177952676
0 1.70e1 17.54
VirtualFunctionEthernet0/a/0-t active 10146115 177952676
0 7.86e1 17.54
arp-input active 1337 1340
0 2.58e3 1.00
dpdk-input polling 6678250972 177954003
0 3.26e3 .03
error-drop active 1324 1327
0 8.68e2 1.00
ethernet-input active 1337 1340
0 1.87e3 1.00
interface-output active 16 16
0 8.06e2 1.00
ip4-icmp-echo-reply active 3 3
0 3.01e4 1.00
ip4-icmp-echo-request active 2 2
0 3.80e3 1.00
ip4-icmp-error active 10146097 177952658
0 1.28e2 17.54
ip4-icmp-input active 5 5
0 2.59e3 1.00
ip4-input-no-checksum active 10146102 177952663
0 4.05e1 17.54
ip4-load-balance active 2 2
0 3.67e3 1.00
ip4-local active 10146102 177952663
0 5.54e1 17.54
ip4-lookup active 20292199 355905321
0 4.37e1 17.54
ip4-rewrite active 10146099 177952660
0 3.23e1 17.54
ip4-udp-lookup active 10146097 177952658
0 3.95e1 17.54
Thanks,
Rahul
On Fri, Aug 11, 2017 at 1:45 AM, Kinsella, Ray <[email protected]>
wrote:
> Hi Rahul,
>
> So there a few additional hoops you have to jump through to get good
> performance. The biggest of which is making sure that
>
> 1. The host is setup to make hugepages available.
> 2. QEMU is pointed at the hugepages.
> 3. The guest is setup to make hugepages available.
>
> The following documented from DPDK covers it in some detail.
>
> http://dpdk.readthedocs.io/en/v16.04/sample_app_ug/vhost.html
>
> Ray K
>
>
> On 09/08/2017 23:33, Rahul Negi wrote:
>
>> Hi All,
>> I am working on an use case where i want to measure the Maximum PPS
>> handled by vpp in sriov configuration.I have created a virtual machine
>> on a host having specifications as follows:-
>> 1.RHEL 7.3 installed
>> 2.Intel X540 10 gig NIC attached
>>
>> I have created a virtual function from one of the interface of 10 gig
>> NIC(for e.g ens3f0) and attached it to my Virtual Machine.I have
>> installed ubuntu 16.04 on my Vm. My vpp version running on vm is 17.10.
>>
>> vpp# show version verbose
>> Version: v17.10-rc0~86-g7d4a22c
>> Compiled by: root
>> Compile host: ubuntu
>> Compile date: Wed Jul 26 18:56:51 EDT 2017
>> Compile location: /root/vpp/vpp
>> Compiler: GCC 5.4.0 20160609
>> Current PID: 5006
>>
>> Currently My Vm is with 4vcpu and vpp cpu model is as follow:-
>> 1.one main thread
>> 2.Three worker threads
>>
>> I am not able to get more than 4Mpps with this configuration of vpp.When
>> i am pumping more than 4Mpps traffic to my VM. I can see this
>> counter rx_no_dma_resources: 5628104 getting incremented at the host.As
>> there is no eth stats available in sriov configuration for virtual
>> functions attached to VM.
>>
>> Guest vcpus are pin to host physical cpus.
>>
>> So 4Mpps is the expected number that we can get in this configuration?
>>
>> Thanks,
>> Rahul
>>
>>
>> _______________________________________________
>> vpp-dev mailing list
>> [email protected]
>> https://lists.fd.io/mailman/listinfo/vpp-dev
>>
>>
_______________________________________________
vpp-dev mailing list
[email protected]
https://lists.fd.io/mailman/listinfo/vpp-dev
