[vpp-dev] hqos configuration fails

[Chidanand]

Hi Everyone,

I wanted to try out QOS  in VPP, for which I am following
https://docs.fd.io/vpp/17.10/qos_doc.html

I have tried the steps  and configs as mentioned in the link above. I am
not able to bring up  the  vpp process.

I am using vpp v22.06

I get following error

vpp[108]: cpu_config: no such thread type 'hqos-threads'

Following is the configuration option I have tried:

cpu {
main-core 1
corelist-workers 5
corelist-hqos-threads 2
}

dpdk {
uio-driver vfio-pci
dev :08:00.0 {
name fpeth1
num-rx-queues 2
num-tx-queues 3
}
dev :09:00.0 {
name fpeth4
num-rx-queues 2
num-tx-queues 3
hqos
}
}

I have isolated cpus on my host and in my VM. I have used the cores  to my
limitations.
Any pointers to resolve this issue will be of great help.

Thanks,
Chidanand

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#22678): https://lists.fd.io/g/vpp-dev/message/22678
Mute This Topic: https://lists.fd.io/mt/97493859/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/1480452/21656/631435203/xyzzy 
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQOS in latest VPP

[Damjan Marion via lists.fd.io]

> On 15.07.2021., at 18:16, satish amara  wrote:
> 
> Hi,
>It looks like Hierarchical Queuing support is not supported in the latest 
> VPP release.  The last release  I see is 20.01.  Any plans to support it 
> again in the future VPP?  I don't see the config commands or nor hqos code in 
> the latest VPP release.

No, currently no plans AFAIK.

— 
Damjan


-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19794): https://lists.fd.io/g/vpp-dev/message/19794
Mute This Topic: https://lists.fd.io/mt/84229266/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] HQOS in latest VPP

[satish amara]

Hi,
   It looks like Hierarchical Queuing support is not supported in the
latest VPP release.  The last release  I see is 20.01.  Any plans to
support it again in the future VPP?  I don't see the config commands or nor
hqos code in the latest VPP release.

Thanks,
Satish K Amara

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#19787): https://lists.fd.io/g/vpp-dev/message/19787
Mute This Topic: https://lists.fd.io/mt/84229266/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] Hqos & vpp 20.09

[Mohsen Meamarian]

Hi
I have noticed that the hqos code in vpp version 20.09 is deprecated
and they are moved to the extras folder inside the src folder. The
hqos command has also been removed from vppctl.
Is there a possibility for qos in vpp 20.09?
If I move the hqos source code into the original vpp code and make the
code, can I use the qos minimum facilities in vpp?

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#18896): https://lists.fd.io/g/vpp-dev/message/18896
Mute This Topic: https://lists.fd.io/mt/81228979/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQOS shape questions #vpp_qos #vpp

[Jasvinder Singh]

May be packet fields (traffic from TRex) associated with pipe, tc, and queues 
are not randomized, so traffic is only going to specific scheduler queues.  
Please refer documentation to know offsets of packet fields.


From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of 
hui12...@163.com
Sent: Friday, September 20, 2019 7:34 AM
To: vpp-dev@lists.fd.io
Subject: [vpp-dev] HQOS shape questions #vpp_qos #vpp


[Edited Message Follows]
I started HQOS, using TREX to send 20Mbps traffic in default configuration. But 
it only can pass 1Mbps. Why?
And I set the subport 0 speed limit of 200 Kbps, it can pass about 666 Kbps, 
why?
Has anyone helped explain why speed limits are ineffective?
Subport 0:
   Rate = 20 bytes/second
   Token bucket size = 100 bytes
   Traffic class rate: TC0 = 20, TC1 = 20, TC2 = 20, TC3 = 20 
bytes/second
   TC period = 10 milliseconds
 Total-Tx:  20.02 Mbps
 Total-Rx: 667.15 Kbps
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#14032): https://lists.fd.io/g/vpp-dev/message/14032
Mute This Topic: https://lists.fd.io/mt/34227753/21656
Mute #vpp: https://lists.fd.io/mk?hashtag=vpp=1480452
Mute #vpp_qos: https://lists.fd.io/mk?hashtag=vpp_qos=1480452
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] HQOS shape questions #vpp_qos #vpp

[hui12723]

[Edited Message Follows]

I started HQOS, using TREX to send 20Mbps traffic in default configuration. But 
it only can pass 1Mbps. Why?
And I set the subport 0 speed limit of 200 Kbps, it can pass about 666 Kbps, 
why?
Has anyone helped explain why speed limits are ineffective?
Subport 0: Rate = 20 bytes/second Token bucket size = 100 bytes Traffic 
class rate: TC0 = 20, TC1 = 20, TC2 = 20, TC3 = 20 bytes/second 
TC period = 10 milliseconds

Total-Tx        :      20.02 Mbps
Total-Rx        :     667.15 Kbps
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#14026): https://lists.fd.io/g/vpp-dev/message/14026
Mute This Topic: https://lists.fd.io/mt/34227753/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] HQOS shape questions #vpp_qos #vpp

[hui12723]

I started HQOS, using TREX to send 20Mbps traffic in default configuration. But 
it only can pass 1Mbps. Why?
And I set the subport 0 speed limit of 200 Kbps, it only can pass about 666 
Kbps, why?
Has anyone helped explain why speed limits are ineffective?
Subport 0: Rate = 20 bytes/second Token bucket size = 100 bytes Traffic 
class rate: TC0 = 20, TC1 = 20, TC2 = 20, TC3 = 20 bytes/second 
TC period = 10 milliseconds

Total-Tx        :      20.02 Mbps
Total-Rx        :     667.15 Kbps
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#14026): https://lists.fd.io/g/vpp-dev/message/14026
Mute This Topic: https://lists.fd.io/mt/34227753/21656
Mute #vpp_qos: https://lists.fd.io/mk?hashtag=vpp_qos=1480452
Mute #vpp: https://lists.fd.io/mk?hashtag=vpp=1480452
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQoS

[Abeeha Aqeel]

Hi Jasvinder,

Yes, I am aware of that. But after applying the patch, assigning hqos-threads 
and enabling hqos on interface the CLI commands don’t show. 

Another issue I have encountered while using vpp version 19.01 is that VPP 
doesn’t start when I assign two corelist-hqos-threads. As shown in the config 
file: 

unix {
    nodaemon
    log /var/log/vpp/vpp.log
    full-coredump
    cli-listen /run/vpp/cli.sock
    # gid vpp
}

api-trace {
  on
}

# api-segment {
    # gid vpp
# }

# socksvr {
    # default
# }

cpu {
    main-core 2
    corelist-workers 4, 20, 6, 22, 8, 24
    corelist-hqos-threads 12,18
}

dpdk {
    dev :05:00.0 {
    num-rx-queues 2
    hqos
    }

    dev :05:00.1 {
    num-rx-queues 2
    # hqos
    }

    # igb_uio
    # uio_pci_generic

    ## Disable multi-segment buffers, improves performance but
    ## disables Jumbo MTU support
    # no-multi-seg

    ## Change hugepages allocation per-socket, needed only if there is need 
for
    ## larger number of mbufs. Default is 256M on each detected CPU socket
    socket-mem 16384, 16384

   ## Disables UDP / TCP TX checksum offload. Typically needed for use
    ## faster vector PMDs (together with no-multi-seg)
    # no-tx-checksum-offload
}

plugins {
    path /root/vpp/build-root/install-vpp_debug-native/vpp/lib/vpp_plugins/
    plugin default { enable }
}
   
I have installed vpp version 19.04 on bare-metal server with centos 7 as 
Operating System and an Intel 10G 2P X520 NIC. What could be the possible issue?


Thank you and regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Singh, Jasvinder
Sent: Wednesday, May 15, 2019 3:19 PM
To: Abeeha Aqeel; Ni, Hongjun
Cc: Byrne, Stephen1; Dumitrescu, Cristian; vpp-dev@lists.fd.io; 
b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

In the config file,  hqos thread is allocated (cpu section), and hqos is’nt 
enabled on interfaces (dpdk section). In the current state, hqos needs separate 
thread (other than worker threads) for scheduling function, and also, 
enablement on dpdk interfaces.

Thanks,
Jasvinder
   

From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com] 
Sent: Wednesday, May 15, 2019 11:03 AM
To: Singh, Jasvinder ; Ni, Hongjun 

Cc: Byrne, Stephen1 ; Dumitrescu, Cristian 
; vpp-dev@lists.fd.io; b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Jasvinder,

Here’s the startup.conf:

unix {
    nodaemon
    log /var/log/vpp/vpp.log
    full-coredump
    cli-listen /run/vpp/cli.sock
    # gid vpp
}

api-trace {
  on
}

# api-segment {
    # gid vpp
# }

# socksvr {
    # default
# }

cpu {
    main-core 2
    corelist-workers 4, 20, 6, 22, 8, 24
    # corelist-hqos-threads 18
}

dpdk {
    dev :05:00.0 {
    num-rx-queues 2
    # hqos
    }

    dev :05:00.1 {
    num-rx-queues 2
    # hqos
    }

    # igb_uio
    # uio_pci_generic

    ## Disable multi-segment buffers, improves performance but
    ## disables Jumbo MTU support
    # no-multi-seg

    ## Change hugepages allocation per-socket, needed only if there is need 
for
    ## larger number of mbufs. Default is 256M on each detected CPU socket
    socket-mem 16384, 16384

   ## Disables UDP / TCP TX checksum offload. Typically needed for use
    ## faster vector PMDs (together with no-multi-seg)
    # no-tx-checksum-offload
}

plugins {
    path /root/vpp/build-root/install-vpp_debug-native/vpp/lib/vpp_plugins/
    plugin default { enable }
}
   

I tried enabling and disabling hqos but it doesn’t work in both cases.

Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Singh, Jasvinder
Sent: Wednesday, May 15, 2019 2:29 PM
To: Abeeha Aqeel; Ni, Hongjun
Cc: Byrne, Stephen1; Dumitrescu, Cristian; vpp-dev@lists.fd.io; 
b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

Can you share startup.conf? 

Thanks,
Jasvinder


From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com] 
Sent: Wednesday, May 15, 2019 7:29 AM
To: Singh, Jasvinder ; Ni, Hongjun 

Cc: Byrne, Stephen1 ; Dumitrescu, Cristian 
; vpp-dev@lists.fd.io; b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Jasvinder, 

The CMakelist.txt files before and after applying the patch are exactly the 
same and hqos plugin is already enabled but the CLI commands are still not 
showing. Below is the screenshot of the CMakelist.text file located at 
/root/vpp/src/plugins/dpdk: 



Thank you and best regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5

Re: [vpp-dev] HQoS

[Jasvinder Singh]

Hi Abeeha,

In the config file,  hqos thread is allocated (cpu section), and hqos is’nt 
enabled on interfaces (dpdk section). In the current state, hqos needs separate 
thread (other than worker threads) for scheduling function, and also, 
enablement on dpdk interfaces.

Thanks,
Jasvinder


From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com]
Sent: Wednesday, May 15, 2019 11:03 AM
To: Singh, Jasvinder ; Ni, Hongjun 

Cc: Byrne, Stephen1 ; Dumitrescu, Cristian 
; vpp-dev@lists.fd.io; b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Jasvinder,

Here’s the startup.conf:

unix {
nodaemon
log /var/log/vpp/vpp.log
full-coredump
cli-listen /run/vpp/cli.sock
# gid vpp
}

api-trace {
  on
}

# api-segment {
# gid vpp
# }

# socksvr {
# default
# }

cpu {
main-core 2
corelist-workers 4, 20, 6, 22, 8, 24
# corelist-hqos-threads 18
}

dpdk {
dev :05:00.0 {
num-rx-queues 2
# hqos
}

dev :05:00.1 {
num-rx-queues 2
# hqos
}

# igb_uio
# uio_pci_generic

## Disable multi-segment buffers, improves performance but
## disables Jumbo MTU support
# no-multi-seg

## Change hugepages allocation per-socket, needed only if there is need 
for
## larger number of mbufs. Default is 256M on each detected CPU socket
socket-mem 16384, 16384

   ## Disables UDP / TCP TX checksum offload. Typically needed for use
## faster vector PMDs (together with no-multi-seg)
# no-tx-checksum-offload
}

plugins {
path /root/vpp/build-root/install-vpp_debug-native/vpp/lib/vpp_plugins/
plugin default { enable }
}


I tried enabling and disabling hqos but it doesn’t work in both cases.

Regards,

Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com<mailto:abeeha.aq...@xflowresearch.com>
www.xflowresearch.com<http://www.xflowresearch.com/>

[cid:6e5d1814-0a3d-444b-9050-060db1465f05]


From: Singh, Jasvinder<mailto:jasvinder.si...@intel.com>
Sent: Wednesday, May 15, 2019 2:29 PM
To: Abeeha Aqeel<mailto:abeeha.aq...@xflowresearch.com>; Ni, 
Hongjun<mailto:hongjun...@intel.com>
Cc: Byrne, Stephen1<mailto:stephen1.by...@intel.com>; Dumitrescu, 
Cristian<mailto:cristian.dumitre...@intel.com>; 
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>; 
b...@xflowresearch.com<mailto:b...@xflowresearch.com>
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

Can you share startup.conf?

Thanks,
Jasvinder


From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com]
Sent: Wednesday, May 15, 2019 7:29 AM
To: Singh, Jasvinder 
mailto:jasvinder.si...@intel.com>>; Ni, Hongjun 
mailto:hongjun...@intel.com>>
Cc: Byrne, Stephen1 
mailto:stephen1.by...@intel.com>>; Dumitrescu, 
Cristian mailto:cristian.dumitre...@intel.com>>; 
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>; 
b...@xflowresearch.com<mailto:b...@xflowresearch.com>
Subject: RE: [vpp-dev] HQoS

Hi Jasvinder,

The CMakelist.txt files before and after applying the patch are exactly the 
same and hqos plugin is already enabled but the CLI commands are still not 
showing. Below is the screenshot of the CMakelist.text file located at 
/root/vpp/src/plugins/dpdk:

[cid:image002.png@01D50B08.CB3C2CA0]

Thank you and best regards,

Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com<mailto:abeeha.aq...@xflowresearch.com>
www.xflowresearch.com<http://www.xflowresearch.com/>

[cid:6e5d1814-0a3d-444b-9050-060db1465f05]


From: Abeeha Aqeel<mailto:abeeha.aq...@xflowresearch.com>
Sent: Wednesday, May 15, 2019 11:04 AM
To: hongjun...@intel.com<mailto:hongjun...@intel.com>
Cc: stephen1.by...@intel.com<mailto:stephen1.by...@intel.com>; 
cristian.dumitre...@intel.com<mailto:cristian.dumitre...@intel.com>; 
cristian.dumitre...@intel.com<mailto:cristian.dumitre...@intel.com>; 
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>; 
b...@xflowresearch.com<mailto:b...@xflowresearch.com>
Subject: FW: [vpp-dev] HQoS


Hi Jasvinder,

The CMakelist.txt files before and after applying the patch are exactly the 
same and hqos plugin is already enabled but the CLI commands are still not 
showing. Below is the screenshot of the CMakelist.text file located at 
/root/vpp/src/plugins/dpdk:

[cid:image002.png@01D50B0D.103CAD30]



Thank you and best regards,

Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com<mailto:abeeha.aq...@xflowresearch.com>
www.xflowresearch.com<http://www.xflowresearch.com/>

[cid:6e5d1814-0a3d-444b-9050-060db1465f05]


From: Singh, Jasvinder<mailto:jasvinder.si...@intel.com>
Sent: Friday, May 10, 2019 2:58 PM
T

Re: [vpp-dev] HQoS

[Abeeha Aqeel]

Hi Jasvinder,

Here’s the startup.conf:

unix {
nodaemon
log /var/log/vpp/vpp.log
full-coredump
cli-listen /run/vpp/cli.sock
# gid vpp
}

api-trace {
  on
}

# api-segment {
# gid vpp
# }

# socksvr {
# default
# }

cpu {
main-core 2
corelist-workers 4, 20, 6, 22, 8, 24
# corelist-hqos-threads 18
}

dpdk {
dev :05:00.0 {
num-rx-queues 2
# hqos
}

dev :05:00.1 {
num-rx-queues 2
# hqos
}

# igb_uio
# uio_pci_generic

## Disable multi-segment buffers, improves performance but
## disables Jumbo MTU support
# no-multi-seg

## Change hugepages allocation per-socket, needed only if there is need 
for
## larger number of mbufs. Default is 256M on each detected CPU socket
socket-mem 16384, 16384

   ## Disables UDP / TCP TX checksum offload. Typically needed for use
## faster vector PMDs (together with no-multi-seg)
# no-tx-checksum-offload
}

plugins {
path /root/vpp/build-root/install-vpp_debug-native/vpp/lib/vpp_plugins/
plugin default { enable }
}
   

I tried enabling and disabling hqos but it doesn’t work in both cases.

Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Singh, Jasvinder
Sent: Wednesday, May 15, 2019 2:29 PM
To: Abeeha Aqeel; Ni, Hongjun
Cc: Byrne, Stephen1; Dumitrescu, Cristian; vpp-dev@lists.fd.io; 
b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

Can you share startup.conf? 

Thanks,
Jasvinder


From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com] 
Sent: Wednesday, May 15, 2019 7:29 AM
To: Singh, Jasvinder ; Ni, Hongjun 

Cc: Byrne, Stephen1 ; Dumitrescu, Cristian 
; vpp-dev@lists.fd.io; b...@xflowresearch.com
Subject: RE: [vpp-dev] HQoS

Hi Jasvinder, 

The CMakelist.txt files before and after applying the patch are exactly the 
same and hqos plugin is already enabled but the CLI commands are still not 
showing. Below is the screenshot of the CMakelist.text file located at 
/root/vpp/src/plugins/dpdk: 



Thank you and best regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Abeeha Aqeel
Sent: Wednesday, May 15, 2019 11:04 AM
To: hongjun...@intel.com
Cc: stephen1.by...@intel.com; cristian.dumitre...@intel.com; 
cristian.dumitre...@intel.com; vpp-dev@lists.fd.io; b...@xflowresearch.com
Subject: FW: [vpp-dev] HQoS


Hi Jasvinder, 

The CMakelist.txt files before and after applying the patch are exactly the 
same and hqos plugin is already enabled but the CLI commands are still not 
showing. Below is the screenshot of the CMakelist.text file located at 
/root/vpp/src/plugins/dpdk: 





Thank you and best regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Singh, Jasvinder
Sent: Friday, May 10, 2019 2:58 PM
To: Ni, Hongjun; Abeeha Aqeel
Cc: Byrne, Stephen1; Dumitrescu, Cristian
Subject: RE: [vpp-dev] HQoS

+ Cristian

From: Singh, Jasvinder 
Sent: Friday, May 10, 2019 10:52 AM
To: Ni, Hongjun ; Abeeha Aqeel 

Cc: Byrne, Stephen1 
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

Looks like HQoS module is disabled in current VPP dpdk plugin (Check 
CMakelist.txt). Please trying enabling it first. You should see HQoS related 
CLIs on the console.

Thanks,
Jasvinder


From: Ni, Hongjun 
Sent: Friday, May 10, 2019 2:47 AM
To: Abeeha Aqeel ; Singh, Jasvinder 

Subject: RE: [vpp-dev] HQoS

Hi Jasvinder,

Could you help to look into this issue?   Thank you!

Thanks,
Hongjun

From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com] 
Sent: Thursday, May 9, 2019 6:02 PM
To: Ni, Hongjun ; Singh, Jasvinder 

Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: RE: [vpp-dev] HQoS

Hi Hongjun, 

I applied the patch submitted to fix the API change on vpp 19.04 using 
underlying dpdk 19.02, following the steps:


git clone https://gerrit.fd.io/r/vppHi Hongjun,

cd vpp
make wipe
make install-dep
make install-ext-deps
wget 
https://gerrit.fd.io/r/changes/16839/revisions/a62db7f3796ef152d23475ed36ad5f3fbfcab2a8/archive?format=tar
mv archive\?format\=tar a62db7f.tar
tar -xvf a62db7f.tar
make build
./vpp -c /root/go/src/github.com/xFlowResearch/BNG/vpp-startup.conf &
./vppctl

I also tried, git fetch and cherry-pick:

cd vpp
make wipe
make install-dep
make install-ext-deps
git stash
git fetch https://gerrit.fd.io/r/vpp refs/changes/39/16839/3 && git cherry-pick 
FETCH_HEAD  
make build
./vpp -c /root/go/src/github.com/xFlowResearch/BNG/vpp-startup.conf &
./vppctl

In both cases, the setup builds and starts properly but hqos plugin doesn’t 
work. Nor does the vpp debug CLI have the hq

FW: [vpp-dev] HQoS

[Abeeha Aqeel]

Hi Jasvinder, 

The CMakelist.txt files before and after applying the patch are exactly the 
same and hqos plugin is already enabled but the CLI commands are still not 
showing. Below is the screenshot of the CMakelist.text file located at 
/root/vpp/src/plugins/dpdk: 





Thank you and best regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Singh, Jasvinder
Sent: Friday, May 10, 2019 2:58 PM
To: Ni, Hongjun; Abeeha Aqeel
Cc: Byrne, Stephen1; Dumitrescu, Cristian
Subject: RE: [vpp-dev] HQoS

+ Cristian

From: Singh, Jasvinder 
Sent: Friday, May 10, 2019 10:52 AM
To: Ni, Hongjun ; Abeeha Aqeel 

Cc: Byrne, Stephen1 
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

Looks like HQoS module is disabled in current VPP dpdk plugin (Check 
CMakelist.txt). Please trying enabling it first. You should see HQoS related 
CLIs on the console.

Thanks,
Jasvinder


From: Ni, Hongjun 
Sent: Friday, May 10, 2019 2:47 AM
To: Abeeha Aqeel ; Singh, Jasvinder 

Subject: RE: [vpp-dev] HQoS

Hi Jasvinder,

Could you help to look into this issue?   Thank you!

Thanks,
Hongjun

From: Abeeha Aqeel [mailto:abeeha.aq...@xflowresearch.com] 
Sent: Thursday, May 9, 2019 6:02 PM
To: Ni, Hongjun ; Singh, Jasvinder 

Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: RE: [vpp-dev] HQoS

Hi Hongjun, 

I applied the patch submitted to fix the API change on vpp 19.04 using 
underlying dpdk 19.02, following the steps:


git clone https://gerrit.fd.io/r/vppHi Hongjun,

cd vpp
make wipe
make install-dep
make install-ext-deps
wget 
https://gerrit.fd.io/r/changes/16839/revisions/a62db7f3796ef152d23475ed36ad5f3fbfcab2a8/archive?format=tar
mv archive\?format\=tar a62db7f.tar
tar -xvf a62db7f.tar
make build
./vpp -c /root/go/src/github.com/xFlowResearch/BNG/vpp-startup.conf &
./vppctl

I also tried, git fetch and cherry-pick:

cd vpp
make wipe
make install-dep
make install-ext-deps
git stash
git fetch https://gerrit.fd.io/r/vpp refs/changes/39/16839/3 && git cherry-pick 
FETCH_HEAD  
make build
./vpp -c /root/go/src/github.com/xFlowResearch/BNG/vpp-startup.conf &
./vppctl

In both cases, the setup builds and starts properly but hqos plugin doesn’t 
work. Nor does the vpp debug CLI have the hqos commands, e.g. in this case it 
should have show dpdk interface hqos placement etc.


I am working on Centos 7 and have assigned 1024 2MB hugepages. I have also 
attached my vpp configuration file with this email. Please suggest what could 
be wrong and why the plugin is not loading properly.



Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Abeeha Aqeel
Sent: Thursday, May 9, 2019 10:10 AM
To: Ni, Hongjun; Singh, Jasvinder
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Hongjun,

Yes, sure. I will try the patch and share my results as soon as possible.


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Ni, Hongjun
Sent: Tuesday, May 7, 2019 6:40 PM
To: Abeeha Aqeel; Singh, Jasvinder
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Abeeha,

There is a patch submitted to fix the API change.   
https://gerrit.fd.io/r/#/c/16839/
Could you help to give it a try?

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 6:32 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Hongjun,

I have tested the HQoS Plugin with iperf3. Below is a simple topology I have 
implemented:
 

First, I tested the default profile (profile 0) provided by VPP and assigned it 
to all 4096 pipes (users) which should give approximately 2MB (10G/4096) to 
each user. It worked fine for 5 users i.e. each user was assigned 2MB but for 
an increased number of users it didn’t work the same way. I changed the token 
bucket rate, token bucket size and tc_period parameters in the hqos.c code 
which didn’t give me any concrete results. I have attached results for 10  
clients with different combinations of token bucket size, token bucket rate and 
tc_period.

Any ideas what the problem might be ? 

Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period  
be assigned? 


S#
Rate (Bytes per second)
Token Bucket Size (Bytes)
TC Period (ms)
Total no of PPPoE sessions
Expected Result (Cumulative average bandwidth in Mbps)
Actual Result (Cumulative average bandwidth in Mbps)
Comments

Subport
Pipe
Subport 
Pipe
Subport
Pipe




1
125000
305175
100
100
10
40
10
24.4
8.1
vpp remains stable
2
125000
305175
100
100
10
60
10
24.4
12.72
vpp remains stable
3
125000
305175
100
100
10
100
10
24.4
11.4
vpp remains stable
4
125000
305175
100
10

Re: [vpp-dev] HQoS

[Abeeha Aqeel]

Hi Hongjun, 

I applied the patch submitted to fix the API change on vpp 19.04 using 
underlying dpdk 19.02, following the steps:


git clone https://gerrit.fd.io/r/vpp
cd vpp
make wipe
make install-dep
make install-ext-deps
wget 
https://gerrit.fd.io/r/changes/16839/revisions/a62db7f3796ef152d23475ed36ad5f3fbfcab2a8/archive?format=tar
mv archive\?format\=tar a62db7f.tar
tar -xvf a62db7f.tar
make build
./vpp -c /root/go/src/github.com/xFlowResearch/BNG/vpp-startup.conf &
./vppctl

I also tried, git fetch and cherry-pick:

cd vpp
make wipe
make install-dep
make install-ext-deps
git stash
git fetch https://gerrit.fd.io/r/vpp refs/changes/39/16839/3 && git cherry-pick 
FETCH_HEAD  
make build
./vpp -c /root/go/src/github.com/xFlowResearch/BNG/vpp-startup.conf &
./vppctl

In both cases, the setup builds and starts properly but hqos plugin doesn’t 
work. Nor does the vpp debug CLI have the hqos commands, e.g. in this case it 
should have show dpdk interface hqos placement etc.


I am working on Centos 7 and have assigned 1024 2MB hugepages. I have also 
attached my vpp configuration file with this email. Please suggest what could 
be wrong and why the plugin is not loading properly.



Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Abeeha Aqeel
Sent: Thursday, May 9, 2019 10:10 AM
To: Ni, Hongjun; Singh, Jasvinder
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Hongjun,

Yes, sure. I will try the patch and share my results as soon as possible.


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Ni, Hongjun
Sent: Tuesday, May 7, 2019 6:40 PM
To: Abeeha Aqeel; Singh, Jasvinder
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Abeeha,

There is a patch submitted to fix the API change.   
https://gerrit.fd.io/r/#/c/16839/
Could you help to give it a try?

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 6:32 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Hongjun,

I have tested the HQoS Plugin with iperf3. Below is a simple topology I have 
implemented:
 

First, I tested the default profile (profile 0) provided by VPP and assigned it 
to all 4096 pipes (users) which should give approximately 2MB (10G/4096) to 
each user. It worked fine for 5 users i.e. each user was assigned 2MB but for 
an increased number of users it didn’t work the same way. I changed the token 
bucket rate, token bucket size and tc_period parameters in the hqos.c code 
which didn’t give me any concrete results. I have attached results for 10  
clients with different combinations of token bucket size, token bucket rate and 
tc_period.

Any ideas what the problem might be ? 

Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period  
be assigned? 


S#
Rate (Bytes per second)
Token Bucket Size (Bytes)
TC Period (ms)
Total no of PPPoE sessions
Expected Result (Cumulative average bandwidth in Mbps)
Actual Result (Cumulative average bandwidth in Mbps)
Comments

Subport
Pipe
Subport 
Pipe
Subport
Pipe




1
125000
305175
100
100
10
40
10
24.4
8.1
vpp remains stable
2
125000
305175
100
100
10
60
10
24.4
12.72
vpp remains stable
3
125000
305175
100
100
10
100
10
24.4
11.4
vpp remains stable
4
125000
305175
100
1000
10
40
10
24.4
12.1
vpp remains stable
5
125000
305175
1000
1000
10
40
10
24.4
12.03
vpp remains stable
6
125000
305175
1000
1000
10
100
10
24.4
2.21
vpp crashes after a few seconds
7
125000
305175
100
100
10
60
10
24.4
12.1
vpp remains stable



Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Ni, Hongjun
Sent: Monday, May 6, 2019 12:52 PM
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

For downstream bandwidth limiting, we leveraged HQos plugin in OpenBRAS 
solution.
In our previous integration test, it could support 64K subscribers with HQos.

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 3:28 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: [vpp-dev] HQoS

Hi Hongjun,  

I have been trying to implement downstream bandwidth limiting using HQoS plugin 
in vpp. It works fine for a certain number of clients (less than 5) but doesn’t 
assign proper bandwidth for larger number of clients. 

Can you please elaborate which method is being used in the OpenBRAS solution 
for band limiting traffic? Are you using the token-bucket algorithm as well?


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (

Re: [vpp-dev] HQoS

[Abeeha Aqeel]

Hi Hongjun,

Yes, sure. I will try the patch and share my results as soon as possible.


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Ni, Hongjun
Sent: Tuesday, May 7, 2019 6:40 PM
To: Abeeha Aqeel; Singh, Jasvinder
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Abeeha,

There is a patch submitted to fix the API change.   
https://gerrit.fd.io/r/#/c/16839/
Could you help to give it a try?

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 6:32 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Hongjun,

I have tested the HQoS Plugin with iperf3. Below is a simple topology I have 
implemented:
 

First, I tested the default profile (profile 0) provided by VPP and assigned it 
to all 4096 pipes (users) which should give approximately 2MB (10G/4096) to 
each user. It worked fine for 5 users i.e. each user was assigned 2MB but for 
an increased number of users it didn’t work the same way. I changed the token 
bucket rate, token bucket size and tc_period parameters in the hqos.c code 
which didn’t give me any concrete results. I have attached results for 10  
clients with different combinations of token bucket size, token bucket rate and 
tc_period.

Any ideas what the problem might be ? 

Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period  
be assigned? 


S#
Rate (Bytes per second)
Token Bucket Size (Bytes)
TC Period (ms)
Total no of PPPoE sessions
Expected Result (Cumulative average bandwidth in Mbps)
Actual Result (Cumulative average bandwidth in Mbps)
Comments

Subport
Pipe
Subport 
Pipe
Subport
Pipe




1
125000
305175
100
100
10
40
10
24.4
8.1
vpp remains stable
2
125000
305175
100
100
10
60
10
24.4
12.72
vpp remains stable
3
125000
305175
100
100
10
100
10
24.4
11.4
vpp remains stable
4
125000
305175
100
1000
10
40
10
24.4
12.1
vpp remains stable
5
125000
305175
1000
1000
10
40
10
24.4
12.03
vpp remains stable
6
125000
305175
1000
1000
10
100
10
24.4
2.21
vpp crashes after a few seconds
7
125000
305175
100
100
10
60
10
24.4
12.1
vpp remains stable



Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Ni, Hongjun
Sent: Monday, May 6, 2019 12:52 PM
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

For downstream bandwidth limiting, we leveraged HQos plugin in OpenBRAS 
solution.
In our previous integration test, it could support 64K subscribers with HQos.

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 3:28 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: [vpp-dev] HQoS

Hi Hongjun,  

I have been trying to implement downstream bandwidth limiting using HQoS plugin 
in vpp. It works fine for a certain number of clients (less than 5) but doesn’t 
assign proper bandwidth for larger number of clients. 

Can you please elaborate which method is being used in the OpenBRAS solution 
for band limiting traffic? Are you using the token-bucket algorithm as well?


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com





-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12961): https://lists.fd.io/g/vpp-dev/message/12961
Mute This Topic: https://lists.fd.io/mt/31517544/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQoS

[Ni, Hongjun]

Hi Abeeha,

There is a patch submitted to fix the API change.
https://gerrit.fd.io/r/#/c/16839/
Could you help to give it a try?

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 6:32 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: Re: [vpp-dev] HQoS

Hi Hongjun,

I have tested the HQoS Plugin with iperf3. Below is a simple topology I have 
implemented:
[cid:image001.png@01D5010B.5C565F10]

First, I tested the default profile (profile 0) provided by VPP and assigned it 
to all 4096 pipes (users) which should give approximately 2MB (10G/4096) to 
each user. It worked fine for 5 users i.e. each user was assigned 2MB but for 
an increased number of users it didn’t work the same way. I changed the token 
bucket rate, token bucket size and tc_period parameters in the hqos.c code 
which didn’t give me any concrete results. I have attached results for 10  
clients with different combinations of token bucket size, token bucket rate and 
tc_period.

Any ideas what the problem might be ?

Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period  
be assigned?


S#

Rate (Bytes per second)

Token Bucket Size (Bytes)

TC Period (ms)

Total no of PPPoE sessions

Expected Result (Cumulative average bandwidth in Mbps)

Actual Result (Cumulative average bandwidth in Mbps)

Comments

Subport

Pipe

Subport

Pipe

Subport

Pipe

1

125000

305175

100

100

10

40

10

24.4

8.1

vpp remains stable

2

125000

305175

100

100

10

60

10

24.4

12.72

vpp remains stable

3

125000

305175

100

100

10

100

10

24.4

11.4

vpp remains stable

4

125000

305175

100

1000

10

40

10

24.4

12.1

vpp remains stable

5

125000

305175

1000

1000

10

40

10

24.4

12.03

vpp remains stable

6

125000

305175

1000

1000

10

100

10

24.4

2.21

vpp crashes after a few seconds

7

125000

305175

100

100

10

60

10

24.4

12.1

vpp remains stable





Regards,

Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com<mailto:abeeha.aq...@xflowresearch.com>
www.xflowresearch.com<http://www.xflowresearch.com/>

[cid:6e5d1814-0a3d-444b-9050-060db1465f05]


From: Ni, Hongjun<mailto:hongjun...@intel.com>
Sent: Monday, May 6, 2019 12:52 PM
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

For downstream bandwidth limiting, we leveraged HQos plugin in OpenBRAS 
solution.
In our previous integration test, it could support 64K subscribers with HQos.

Thanks,
Hongjun

From: vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io> 
[mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 3:28 PM
To: Ni, Hongjun mailto:hongjun...@intel.com>>
Cc: b...@xflowresearch.com<mailto:b...@xflowresearch.com>; 
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>
Subject: [vpp-dev] HQoS

Hi Hongjun,

I have been trying to implement downstream bandwidth limiting using HQoS plugin 
in vpp. It works fine for a certain number of clients (less than 5) but doesn’t 
assign proper bandwidth for larger number of clients.

Can you please elaborate which method is being used in the OpenBRAS solution 
for band limiting traffic? Are you using the token-bucket algorithm as well?


Regards,

Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com<mailto:abeeha.aq...@xflowresearch.com>
www.xflowresearch.com<http://www.xflowresearch.com/>

[cid:6e5d1814-0a3d-444b-9050-060db1465f05]



-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12950): https://lists.fd.io/g/vpp-dev/message/12950
Mute This Topic: https://lists.fd.io/mt/31517544/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQoS

[Abeeha Aqeel]

Hi Hongjun,

I have tested the HQoS Plugin with iperf3. Below is a simple topology I have 
implemented:
 

First, I tested the default profile (profile 0) provided by VPP and assigned it 
to all 4096 pipes (users) which should give approximately 2MB (10G/4096) to 
each user. It worked fine for 5 users i.e. each user was assigned 2MB but for 
an increased number of users it didn’t work the same way. I changed the token 
bucket rate, token bucket size and tc_period parameters in the hqos.c code 
which didn’t give me any concrete results. I have attached results for 10  
clients with different combinations of token bucket size, token bucket rate and 
tc_period.

Any ideas what the problem might be ? 

Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period  
be assigned? 


S#
Rate (Bytes per second)
Token Bucket Size (Bytes)
TC Period (ms)
Total no of PPPoE sessions
Expected Result (Cumulative average bandwidth in Mbps)
Actual Result (Cumulative average bandwidth in Mbps)
Comments

Subport
Pipe
Subport 
Pipe
Subport
Pipe




1
125000
305175
100
100
10
40
10
24.4
8.1
vpp remains stable
2
125000
305175
100
100
10
60
10
24.4
12.72
vpp remains stable
3
125000
305175
100
100
10
100
10
24.4
11.4
vpp remains stable
4
125000
305175
100
1000
10
40
10
24.4
12.1
vpp remains stable
5
125000
305175
1000
1000
10
40
10
24.4
12.03
vpp remains stable
6
125000
305175
1000
1000
10
100
10
24.4
2.21
vpp crashes after a few seconds
7
125000
305175
100
100
10
60
10
24.4
12.1
vpp remains stable



Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



From: Ni, Hongjun
Sent: Monday, May 6, 2019 12:52 PM
Subject: RE: [vpp-dev] HQoS

Hi Abeeha,

For downstream bandwidth limiting, we leveraged HQos plugin in OpenBRAS 
solution.
In our previous integration test, it could support 64K subscribers with HQos.

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 3:28 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: [vpp-dev] HQoS

Hi Hongjun,  

I have been trying to implement downstream bandwidth limiting using HQoS plugin 
in vpp. It works fine for a certain number of clients (less than 5) but doesn’t 
assign proper bandwidth for larger number of clients. 

Can you please elaborate which method is being used in the OpenBRAS solution 
for band limiting traffic? Are you using the token-bucket algorithm as well?


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com




-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12934): https://lists.fd.io/g/vpp-dev/message/12934
Mute This Topic: https://lists.fd.io/mt/31517544/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQoS

[Ni, Hongjun]

Hi Abeeha,

For downstream bandwidth limiting, we leveraged HQos plugin in OpenBRAS 
solution.
In our previous integration test, it could support 64K subscribers with HQos.

Thanks,
Hongjun

From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Abeeha Aqeel
Sent: Monday, May 6, 2019 3:28 PM
To: Ni, Hongjun 
Cc: b...@xflowresearch.com; vpp-dev@lists.fd.io
Subject: [vpp-dev] HQoS

Hi Hongjun,

I have been trying to implement downstream bandwidth limiting using HQoS plugin 
in vpp. It works fine for a certain number of clients (less than 5) but doesn’t 
assign proper bandwidth for larger number of clients.

Can you please elaborate which method is being used in the OpenBRAS solution 
for band limiting traffic? Are you using the token-bucket algorithm as well?


Regards,

Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com<mailto:abeeha.aq...@xflowresearch.com>
www.xflowresearch.com<http://www.xflowresearch.com/>

[cid:6e5d1814-0a3d-444b-9050-060db1465f05]


-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12931): https://lists.fd.io/g/vpp-dev/message/12931
Mute This Topic: https://lists.fd.io/mt/31517544/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] HQoS

[Abeeha Aqeel]

Hi Hongjun,  

I have been trying to implement downstream bandwidth limiting using HQoS plugin 
in vpp. It works fine for a certain number of clients (less than 5) but doesn’t 
assign proper bandwidth for larger number of clients. 

Can you please elaborate which method is being used in the OpenBRAS solution 
for band limiting traffic? Are you using the token-bucket algorithm as well?


Regards,
 
Abeeha Aqeel
Network Design Engineer
Xflow Research Inc.
+923245062309 (GMT+5)
abeeha.aq...@xflowresearch.com
www.xflowresearch.com



-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12930): https://lists.fd.io/g/vpp-dev/message/12930
Mute This Topic: https://lists.fd.io/mt/31517544/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQoS- Traffic shaping - Pipe profiles

[Damjan Marion via Lists.Fd.Io]

Dear Abeeha,

I’m afraid you will not get much luck here. DPDK HQoS support code was 
contributed long time ago, it is incomplete 
and not actively developed. Also seems that it is not used as i disabled it in 
master 3 months ago (due to some API changes)
and I didn’t hear any complaint since.

Personally I’m very keen to remove it completely unless somebody is keen to 
pick it up and actively work on it.

— 
Damjan


> On 2 May 2019, at 14:47, Abeeha Aqeel  wrote:
> 
> Hi everyone, 
>  
> I’ve been testing the behavior of the HQoS plugin in VPP. As for now there is 
> only a single subport (Subport 0),  and 4096 pipes in each subport. I tested 
> HQoS for the default profile (profile 0) which should assign approximately 
> 2-3Mbps to each user on a 10Gbps link. For two iperf3 clients simultaneously 
> sending traffic to vpp, HQoS is working fine and limiting traffic to 2Mbps 
> per client but for greater number of clients it does not function the same 
> way.
>  
> Below is a simple topology I have implemented to test the HQoS functionality. 
> I tried changing the token bucket rate, token bucket size and tc_period 
> values but still not much luck. I have attached results for 10  clients with 
> different combinations of token bucket size, token bucket rate and tc_period.
>  
> Any ideas what the problem might be ?
>  
> Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period 
> should be set ?
> <0216E4862CAB4884A673627A3EB8B833.png> 
>  
> S#
> Rate (Bytes per second)
> Token Bucket Size (Bytes)
> TC Period (ms)
> Total no of PPPoE sessions
> Expected Result (Commulative average bandwidth in Mbps)
> Actual Result (Commulative average bandwidth in Mbps)
> Comments
> Subport
> Pipe
> Subport 
> Pipe
> Subport
> Pipe
> 1
> 125000
> 305175
> 100
> 100
> 10
> 40
> 10
> 24.4
> 8.1
> vpp remains stable
> 2
> 125000
> 305175
> 100
> 100
> 10
> 60
> 10
> 24.4
> 12.72
> vpp remains stable
> 3
> 125000
> 305175
> 100
> 100
> 10
> 100
> 10
> 24.4
> 11.4
> vpp remains stable
> 4
> 125000
> 305175
> 100
> 1000
> 10
> 40
> 10
> 24.4
> 12.1
> vpp remains stable
> 5
> 125000
> 305175
> 1000
> 1000
> 10
> 40
> 10
> 24.4
> 12.03
> vpp remains stable
> 6
> 125000
> 305175
> 1000
> 1000
> 10
> 100
> 10
> 24.4
> 2.21
> vpp crashes after a few seconds
> 7
> 125000
> 305175
> 100
> 100
> 10
> 60
> 10
> 24.4
> 12.1
> vpp remains stable
>  
>  
>  
>  
> Regards,
>  
> Abeeha Aqeel
> Network Design Engineer, 
> xFlow Research Inc. 
> abeeha.aq...@xflowresearch.com 
>  
> <0216E4862CAB4884A673627A3EB8B833.png>-=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> 
> View/Reply Online (#12897): https://lists.fd.io/g/vpp-dev/message/12897 
> 
> Mute This Topic: https://lists.fd.io/mt/31454199/675642 
> 
> Group Owner: vpp-dev+ow...@lists.fd.io 
> Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub 
>   [dmar...@me.com 
> ]
> -=-=-=-=-=-=-=-=-=-=-=-

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12916): https://lists.fd.io/g/vpp-dev/message/12916
Mute This Topic: https://lists.fd.io/mt/31454199/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] HQoS- Traffic shaping - Pipe profiles

[Abeeha Aqeel]

Hi everyone, 

I’ve been testing the behavior of the HQoS plugin in VPP. As for now there is 
only a single subport (Subport 0),  and 4096 pipes in each subport. I tested 
HQoS for the default profile (profile 0) which should assign approximately 
2-3Mbps to each user on a 10Gbps link. For two iperf3 clients simultaneously 
sending traffic to vpp, HQoS is working fine and limiting traffic to 2Mbps per 
client but for greater number of clients it does not function the same way. 

Below is a simple topology I have implemented to test the HQoS functionality. I 
tried changing the token bucket rate, token bucket size and tc_period values 
but still not much luck. I have attached results for 10  clients with different 
combinations of token bucket size, token bucket rate and tc_period.

Any ideas what the problem might be ? 

Also, in what proportion should the tb_rate, tb_size , tc_rate and tc_period 
should be set ? 
 

S#
Rate (Bytes per second)
Token Bucket Size (Bytes)
TC Period (ms)
Total no of PPPoE sessions
Expected Result (Commulative average bandwidth in Mbps)
Actual Result (Commulative average bandwidth in Mbps)
Comments

Subport
Pipe
Subport 
Pipe
Subport
Pipe




1
125000
305175
100
100
10
40
10
24.4
8.1
vpp remains stable
2
125000
305175
100
100
10
60
10
24.4
12.72
vpp remains stable
3
125000
305175
100
100
10
100
10
24.4
11.4
vpp remains stable
4
125000
305175
100
1000
10
40
10
24.4
12.1
vpp remains stable
5
125000
305175
1000
1000
10
40
10
24.4
12.03
vpp remains stable
6
125000
305175
1000
1000
10
100
10
24.4
2.21
vpp crashes after a few seconds
7
125000
305175
100
100
10
60
10
24.4
12.1
vpp remains stable



Regards,
 
Abeeha Aqeel
Network Design Engineer, 
xFlow Research Inc. 
abeeha.aq...@xflowresearch.com

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12897): https://lists.fd.io/g/vpp-dev/message/12897
Mute This Topic: https://lists.fd.io/mt/31454199/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] hqos: multicore and multi-network interface problem

[avakili]

Hi dears,
QoS using hqos feature is working with single network interface and single core 
with following config:

##IT WORKS##
dpdk {
   dev :02:00.0 {
 num-rx-queues 2
 hqos
   }
    }
   
    cpu {
  main-core 0
  corelist-workers  1, 2, 3, 4
  corelist-hqos-threads  5

But when using more than one interface or core then not working:

##NOT WORKING##
dpdk { 
   dev :02:00.0 {
 num-rx-queues 2
 hqos
   }
   dev :06:00.0 {
 num-rx-queues 2
    hqos
  }
    }
   
    cpu {
  main-core 0
  corelist-workers  1, 2, 3, 4
  corelist-hqos-threads  5, 6
    }

Please help to fix or any idea if it is a bug in VPP.
Thanks
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10776): https://lists.fd.io/g/vpp-dev/message/10776
Mute This Topic: https://lists.fd.io/mt/27120560/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] hqos configuration in vpp (Connection refused)

[Alex]

The same problem with you. Have you been solved the problem?
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10483): https://lists.fd.io/g/vpp-dev/message/10483
Mute This Topic: https://lists.fd.io/mt/25396306/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] hqos configuration in vpp (Connection refused)

[Jasvinder Singh]

Hi Reza,

I tested HQoS on my fedora system as below and found it working, didn’t see any 
issue;


1.   Make build

2.   Make run STARTUP_CONF=startup.conf

3.   DGBvpp#

startup.conf

unix {
interactive
log /tmp/vpp.log
cli-listen /run/vpp/cli.sock
full-coredump
}

dpdk {
socket-mem 8192,8192

dev :02:00.0 {num-rx-queues 2 hqos}
dev :02:00.1 {num-rx-queues 2 hqos}

num-mbufs 100

}

cpu {
main-core 0
corelist-workers 1,2,3,4
corelist-hqos-threads  5,6
}

Thanks,
Jasvinder


From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Reza Mirzaei
Sent: Sunday, September 9, 2018 9:11 AM
To: Vpp Dev 
Subject: [vpp-dev] hqos configuration in vpp (Connection refused)


Hi

I want to config vpp in a way that it can use hqos according to this 
tutorial<https://docs.fd.io/vpp/16.12/qos_doc.html>, but when i set hqos 
feature for an interface in /etc/vpp/starttup.conf and then restart vpp 
service, it's not clear for me that why vpp service can not be started. can 
anyone help me to solve this problem?

My modifications on /etc/vpp/startup.conf are as follows:



 ...

 cpu {

 main-core 0
 corelist-workers 1, 2
 corelist-hqos-threads 3
 }

 dpdk {

 socket-mem 2048,2048

 dev :0b:00.0 {
 num-rx-queues 2 hqos
 }

 num-mbufs 100

 }

...



Output of vppctl command after service vpp restart:

   clib_socket_init: connect (fd 3, '/run/vpp/cli.sock'): Connection 
refused



Best regards

Reza
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10482): https://lists.fd.io/g/vpp-dev/message/10482
Mute This Topic: https://lists.fd.io/mt/25396306/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] hqos configuration in vpp (Connection refused)

[Reza Mirzaei]

Hi 

I want to config vpp in a way that it can use hqos according to this
tutorial [1], but when i set hqos feature for an interface in
/etc/vpp/starttup.conf and then restart vpp service, it's not clear for
me that why vpp service can not be started. can anyone help me to solve
this problem? 

My modifications on /etc/vpp/startup.conf are as follows: 

 ... 

 cpu { 

 main-core 0
  corelist-workers 1, 2
  corelist-hqos-threads 3
 } 

 dpdk { 

 socket-mem 2048,2048 

 dev :0b:00.0 {
  num-rx-queues 2 hqos
  } 

 num-mbufs 100 

 } 

... 

Output of _VPPCTL_ command after _SERVICE VPP RESTART_: 

   clib_socket_init: connect (fd 3, '/run/vpp/cli.sock'):
Connection refused 

Best regards 

Reza 

Links:
--
[1] https://docs.fd.io/vpp/16.12/qos_doc.html
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10447): https://lists.fd.io/g/vpp-dev/message/10447
Mute This Topic: https://lists.fd.io/mt/25396306/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] HQoS pipe profiles

[David Hotham via vpp-dev]

Thanks - that's exactly what I'm doing, so it's good to hear that it's sensible

From: Singh, Jasvinder [mailto:jasvinder.si...@intel.com]
Sent: 13 October 2016 11:04
To: David Hotham
Cc: vpp-dev@lists.fd.io
Subject: RE: HQoS pipe profiles

To avoid frequent changes in source code and compilations, you can add number 
of pipe profiles in the source code (right now, there is only profile and all 
pipes uses that profile). In run-time, you can assign new pipe profile to the 
any pipe using following CLI;
set dpdk interface hqos pipe  subport  pipe  profile 
or using binary API
sw_interface_set_dpdk_hqos_pipe rx  | sw_if_index  subport 
 pipe  profile 

From: vpp-dev-boun...@lists.fd.io<mailto:vpp-dev-boun...@lists.fd.io> 
[mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Singh, Jasvinder
Sent: Wednesday, October 12, 2016 9:25 PM
To: David Hotham 
<david.hot...@metaswitch.com<mailto:david.hot...@metaswitch.com>>
Cc: vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>
Subject: Re: [vpp-dev] HQoS pipe profiles

Hi David,

Currently, there is no cli or binary API for changing the pipe profile.

Thanks,
Jasvinder


From: vpp-dev-boun...@lists.fd.io<mailto:vpp-dev-boun...@lists.fd.io> 
[mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of David Hotham via vpp-dev
Sent: Wednesday, October 12, 2016 5:47 PM
To: vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>
Subject: [vpp-dev] HQoS pipe profiles

Am I right in thinking that currently the only way to configure pipe profiles 
is via the source code and recompiling?
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev