From mboxrd@z Thu Jan  1 00:00:00 1970
Content-Type: text/plain; charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\))
From: arisawa <arisawa@ar.aichi-u.ac.jp>
In-Reply-To: <CAJSxfmLXKBwqi1NuAqPvT9T8ZOqGYwxqAz0p=_u=9+PuEe5deQ@mail.gmail.com>
Date: Thu, 19 May 2016 13:07:27 +0900
Content-Transfer-Encoding: quoted-printable
Message-Id: <23B73D86-C256-47D2-8329-8C42E4FFA77E@ar.aichi-u.ac.jp>
References: <DAF4E6F1-91EB-46EE-A844-A9A3AE812B8B@ar.aichi-u.ac.jp>
	<CAOw7k5hQJD1Qo9GoAzNFzrqD0VYXr+yzxRnDam3_UDGPBvcOXg@mail.gmail.com>
	<CAJSxfmLXKBwqi1NuAqPvT9T8ZOqGYwxqAz0p=_u=9+PuEe5deQ@mail.gmail.com>
To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net>
Subject: Re: [9fans] bug in authdial()
Topicbox-Message-UUID: 8fe010de-ead9-11e9-9d60-3106f5b1d025

hello,

thanks for the information.
the lookup depends on dns cache but I also don=E2=80=99t think it has a =
problem.
I think the negotiation process is innocent.
plan9 auth negotiation is fairly simple, that needs only a single step:
	C=E2=86=92A:  ticket request
	A=E2=86=92C:  encrypted ticket
I measured the times that are required for the auth server to complete =
the ticket request.
they are only a few milliseconds.

dialing time from client  to auth server is the problem.


> 2016/05/19 2:03=E3=80=81Skip Tavakkolian <skip.tavakkolian@gmail.com> =
=E3=81=AE=E3=83=A1=E3=83=BC=E3=83=AB=EF=BC=9A
>=20
> i don't think it's the auth lookup:
>=20
> supermic% time ndb/csquery /net/cs net!luna.nyx.link!ticket
> /net/tcp/clone 115.36.102.252!567
> /net/tcp/clone 2402:6b00:4040:b600::8!567
> 0.00u 0.00s 0.34r 	 ndb/csquery /net/cs net!luna.nyx.link!ticket
> supermic% time ndb/csquery /net/cs net!grid.nyx.link!cpu
> /net/tcp/clone 115.36.102.252!17010
> /net/tcp/clone 2402:6b00:4040:b600::9!17010
> 0.00u 0.00s 0.01r 	 ndb/csquery /net/cs net!grid.nyx.link!cpu
> supermic%=20
>=20
> i suspect it's auth negotiations.
>=20
>=20
>=20
> On Wed, May 18, 2016 at 6:07 AM Charles Forsyth =
<charles.forsyth@gmail.com> wrote:
>=20
> On 18 May 2016 at 01:43, arisawa <arisawa@ar.aichi-u.ac.jp> wrote:
> -                       p =3D netmkaddr(nt->val, netroot, "ticket");
>                         rv =3D dial(p, 0, 0, 0);
>                         if(rv >=3D 0)
>                                 break;
>=20
> +                       p =3D netmkaddr(nt->val, =E2=80=9Ctcp", =
"ticket");
>                         rv =3D dial(p, 0, 0, 0);
>                         if(rv >=3D 0)
>                                 break;
> +                       p =3D netmkaddr(nt->val, =E2=80=9Cil", =
"ticket");
> +                       rv =3D dial(p, 0, 0, 0);
> +                       if(rv >=3D 0)
> +                               break;
>=20
> But that's just (eventually) moving the cs search into every =
application and bound to specific network types.
> Why is the cs search with "net" so slow?