Re: [vpp-dev] Bgp router implementation

BGP functional in VPP is good solution, vote for this offer. Also will be great to have traffic shaper in vpp stack.

Re: [vpp-dev] VPP RPMs available in Centos

Jerome, Currently, it is a manual process but not a terribly time intensive one. 18.04 and 18.01.1 are in the mirrors now. 18.01.2 within a week. 18.07 will require a minor bug fix in Centos infra with anticipated for Centos 7.5 hopefully before vpp 18.07 release date. --Tom On

Re: [EXT] [vpp-dev] Marvell's AI from last call

Damjan is right. It does not make sense to over wrapup (overheads) when it is not needed.

Re: [vpp-dev] VLIB_BUFFER_TOTAL_LENGTH_VALID set but invalid total_length_not_including_first_buffer

Hello Florin, Yes, that would do the job :) So, with your suggestion, the semantic of total_length_not_including_first_buffer is, that it is correct if and only if (NEXT_PRESENT & TOTAL_LENGTH_VALID), rather than if (TOTAL_LENGTH_VALID) as one might think at the beginning. It seems to make

Re: [vpp-dev] query on use_pthread in startup.conf

You cannot do that if you use DPDK. If you disable DPDK you will get all worker threads started as pthreads. And to be even more precise, DPDK threads are also pthreads, they just come with some extra stuff. What do you want to achieve? — Damjan > On 17 May 2018, at 13:43, bindiya Kurle

Re: [EXT] [vpp-dev] Marvell's AI from last call

It is faster because there is no DPDK overhead — Damjan > On 23 May 2018, at 17:45, Tina Tsou wrote: > > Dear Damjan, > > Is the 30% faster because of using different applications or something else? > > > Thank you, > Tina > > On May 23, 2018, at 7:56 AM, Damjan

Re: [EXT] [vpp-dev] Marvell's AI from last call

Dear Damjan, Is the 30% faster because of using different applications or something else? Thank you, Tina On May 23, 2018, at 7:56 AM, Damjan Marion > wrote: I addef -2 on it as there is no value in having it in vpp. We already have

Re: [vpp-dev] VLIB_BUFFER_TOTAL_LENGTH_VALID set but invalid total_length_not_including_first_buffer

We should simply deprecate that flag and put expectation on device driver including dpdk-input to always set total_length_not_including_first_buffer in case when buffers are chained (VLIB_BUFER_NEXT_VALID is set). — Damjan > On 23 May 2018, at 17:13, Yoann Desmouceaux

Re: [vpp-dev] VLIB_BUFFER_TOTAL_LENGTH_VALID set but invalid total_length_not_including_first_buffer

Hi Yoann, Shouldn’t the clone code only read s->total_length_not_including_first_buffer only if VLIB_BUFFER_NEXT_PRESENT flag is set for s? I’m thinking about this line in particular: d->total_length_not_including_first_buffer = s->total_length_not_including_first_buffer +

[vpp-dev] VLIB_BUFFER_TOTAL_LENGTH_VALID set but invalid total_length_not_including_first_buffer

Hello, I noticed that dpdk devices set the VLIB_BUFFER_TOTAL_LENGTH_VALID flag when receiving packets, regardless of the value of total_length_not_including_first_buffer. Since dpdk only copies one cacheline from its vlib_buffer template into the buffer upon receipt of a packet,

Re: [EXT] [vpp-dev] Marvell's AI from last call

I addef -2 on it as there is no value in having it in vpp. We already have native driver which is 30% faster. At the end dpdk driver is just a wrapper around musdk — Damjan > On 22 May 2018, at 20:05, Szymon Śliwa wrote: > > Hi, I improved the patch Natalie submitted,

Re: [vpp-dev] VPP RPMs available in Centos

Hey Thomas, Can you let us know what you are planning to do for next releases? Are you manually building those RPMs or will you automatically include bugfix versions as well as future versions (e.g. 18.07)? Jerome On 5/21/2018 12:24 PM, Thomas F Herbert wrote: VPP 18.04 RPMs are available in

[vpp-dev] Experimental: MDR Search

Hi, We have been working on improving NDR/PDR search for CSIT throughput tests, focusing specifically on reducing the overall search time so that we could run them more frequently. The result is a new duration aware multi-phase multi-rate search algorithm, that we call Multiple Drop Rate (MDR).

Re: [vpp-dev] VPP RPMs available in Centos

W00t Great news folks!! Ray K From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Jerome Tollet Sent: Tuesday 22 May 2018 09:18 To: Dave Wallace ; vpp-dev@lists.fd.io Subject: Re: [vpp-dev] VPP RPMs available in Centos Excellent! Thanks.