Discussion:
Porting lagg(4) from FreeBSD
(too old to reply)
s ymgch
2021-04-02 04:40:46 UTC
Permalink
Hi tech-net@,

I am porting lagg(4) from FreeBSD for making the feature of link-aggregation
mp-safe and for improving functions.
I put the patch at following links, and I'd like your comment about it.

The patch is here:
http://netbsd.org/~yamaguchi/lagg.patch
and
https://github.com/s-ymgch228/netbsd-src_offficial/pull/1/files

Thanks,
--
yamaguchi

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
s ymgch
2021-04-05 02:00:11 UTC
Permalink
(added CC: tech-net@)
Hi,

Yes, your guess is correct. And my main motivation is also
to make agr(4) mp-scalable.
I tried to change agr(4) before the porting, but I gave up
because agr(4) has some bugs and it is a complicated implementation.

In addition to this reason, lagg(4) supports methods to create
a Link-Aggregation-Group(LAG) named "failover" and "loadbalance" besides
"lacp" that agr(4) also supports. And I want to use them.
--
yamaguchi
Hi, Yamaguchi-san!
Can you say anything about how it compares or relates to agr(4)?
(I haven't looked at lagg(4) at all yet, and I don't know much about
agr(4); this is not criticism of the work you did! Just curious how
it relates to the other thing we have with `link aggregation' in the
title. I'm guessing that agr(4) is not MP-scalable or something.)
Date: Fri, 2 Apr 2021 13:40:46 +0900
I am porting lagg(4) from FreeBSD for making the feature of link-aggregation
mp-safe and for improving functions.
I put the patch at following links, and I'd like your comment about it.
http://netbsd.org/~yamaguchi/lagg.patch
and
https://github.com/s-ymgch228/netbsd-src_offficial/pull/1/files
Thanks,
--
yamaguchi
--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Christos Zoulas
2021-04-05 13:28:03 UTC
Permalink
Post by s ymgch
Hi,
Yes, your guess is correct. And my main motivation is also
to make agr(4) mp-scalable.
I tried to change agr(4) before the porting, but I gave up
because agr(4) has some bugs and it is a complicated implementation.
In addition to this reason, lagg(4) supports methods to create
a Link-Aggregation-Group(LAG) named "failover" and "loadbalance" besides
"lacp" that agr(4) also supports. And I want to use them.
We should remove agr eventually though..

christos


--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
s ymgch
2021-04-20 05:03:40 UTC
Permalink
Hi,

Do you have any questions or comments?
I would like to commit this about the middle of May if there is nothing wrong.
Post by s ymgch
I am porting lagg(4) from FreeBSD for making the feature of link-aggregation
mp-safe and for improving functions.
I put the patch at following links, and I'd like your comment about it.
http://netbsd.org/~yamaguchi/lagg.patch
and
https://github.com/s-ymgch228/netbsd-src_offficial/pull/1/files
Thanks,
--
yamaguchi
--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Loading...