IETF onsite networks; discussion, cash, knowledge

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Mr. Crocker & Mr. Hain,

I use this message only because it was available, and it somewhat sums up
the assumptions and conclusions reached previously many times.  I use it
as an outline only.  Please do not view this as any kind of comment on you
or your thoughts, words or expressions directly.

To the IETF @ Large;

>  On Thu, 17 Mar 2005 07:51:27 -0800, Tony Hain wrote:
>>  The fact that
Actually are considered 'emergency' at this point shows that in fact
people
>> do
>>  expect new features from the equipment,
>
>
> it shows that SOME people decided to add new features.

Actaully not.  It shows that new features came about because of the
available hardware.  Why deploy 15 Cisco 340's that all need firmware
upgrades when you get an offer by "Company X" to give you new, easier to
manage hardware, that has added value that will make it easier to find
problems before IETF-ers-at-large start grumbling under their breath about
it?  Maybe you have an answer.  But have you been on even one of the NOC
teams for IETF 58 through 62?  If no -> Mod down 20 points, then
re-evaluate.

> it does not show that the network really needed them.

Scenario:  Ad-Hoc networks.
discussion
Before:  There is an ad-hoc network somewhere on the "x" floor.  They are
sucking people into them.  People are complaining.  Person from company X
walks by:  "Hey I have a product that will pin point that person so you
can show them the light"

After:  Hmm, not as many ad-hoc networks.  We can exclude those people
from the net so we isolate them.  Extra features that come along for the
ride, maybe not as proven as previous, but those new features do help the
ALL VOLUNTEER NOC TEAM to isolate some of those problems.

> and the difference between these two points is exactly where open
discussion and agreement among the ietf meeting participants could be
helpful.

Open discussion would be great.  However, on this list, probably not the
best.  I think IETF people are great, and very smart people.  But from
watching the re-org procedure, there are many want-to-be lawyers,
accountants, business executives, etc.  I'm sure there were many people
commenting on those topics that actually were official members of those
professions.  The point I am making is that running an IETF NOC is so
different that at some point in the interest of the signal to noise ratio
we have to let the bodies in the trenches fight the war and come to that
table with their comments and quite possibly tone down the comments of
IETF meeting participants.  While their comments should be welcome, they
simply don't have the view from both sides of the switches & APs.

I'm all for dicussion, but if your name hasn't been on at least one thank
you list for "NOC team IETF now minus five inclusive" then chances are
your opinions are welcome but possibly not up to date.

Scenario:  Wired drops for the presenter, jabber scribe, etc.
This would be great.  However, this of course adds hardware.  Quite a bit
of it.  I put a switch in a central like location and secure it, POE the
AP's and run 4-6 AP's in multiple rooms with 4-6 CAT5 cables.  Now, I need
to put breakout switches in each room, secure them, tape down the cables,
possibly pull the cables and put them back down if the meeting rooms get
used by other groups.  I need to ship the cables, and the extra switches,
along with the security cables.  These issues continue.

Result:  This is no reason that this cannot be accomplished.  Why is it
difficult?  Security, time to deploy & tape down cat5 cable, shipping and
acquiring extra hardware.

Consequence:  It is easy to say, this should be done.  It's great to talk
about it and come up with the 7 ways it can be accomplished.  It would be
perfect if you could pay for it too.  Until then, you are back to getting
what you pay for.

>
> right now, the folks doing the choosing pretty much have to guess what
the
> folks doing the using want/need.  open discussion could eliminate the
guessing.

Choosing would be an excellent gift to have when doing a non-hosted
terminal room.  When there isn't a host, you take whatever you can get. 
When I did my first terminal room in San Francisco, Cisco stepped up to
the plate with 2 pallets of gear.  I still had to worry about: Terminals,
wiring, placement, extra routers, extra switches, fiber, this list
continues ad nauseum.  Choice you say?  My choice was either Cisco (which
worked fabulously I might add) or to be incredibly delusional and wait for
something else.

> if, as some have voiced, the community of attendees feel it is essential
that we eat our own dogfood of new features, on our meeting production
network, then we will have agreed to the consequence, assured lack of
stability.

> if instead the community feels that reliability for a core set of
functions is paramount, then new features can only be added after they
are
> viewed as stable and low-risk.

Which all goes out the window when you settle for the "You get what you
don't pay for" scenario.  Unless you a throwing money at the problem, you
don't get to dictate which solution to chose.  Please, PLEASE don't quote
the meeting fees, etc. at me.  I've done the budgets, I've seen the
costs.Internet
 I've worked for that company.  I've run that NOC.  You take what you can
get.  Even then in some instances, you get what you can take.

Are you paying for the hardware at a hosted IETF?  No.
Are you paying for the hardware at a non-hosted IETF?  No.
Are you getting what you pay for?  Yes, and A LOT MORE!
consistencypersistence
In Summary:  The fees that attendees pay don't come CLOSE to paying for
what is required to run a terminal room.  People, product and sanity.  The
numbers to design/engineer, create, ship, deploy, test, maintain, run and
take down a NOC are not fictious unknown numbers, these can be calculated.
 I've done it before.

No host?  No Sponsor?  Volunteers?  Keep in mind that except for me, the
people who ran IETF NOCS for un-hosted meetings were there to actually
attend meetings.  I can tell you, they didn't make too many of them.

Hardware?  You take what you can get.  If someone wants to donate
hardware, who do they donate it to?  Foretec?  CNRI?  The IETF doesn't
exist as an entity to donate to (at least that's the line I got when I
asked that question).  Then when it breaks, who fixes, it?  Who pays for
the flight cases to ship the stuff so it won't get destroyed?  If hosts
ask to use it, do we let them?  Who pays for customs on that?

Donating hardware or sponsoring.  I've seen a couple of IETFs that I think
went pretty well.  IETF Austria and IETF Yokohama (this list is larger,
but let's use these two as an example).  I saw that Austria had a great
team of employees working on the network for over a year, and I also saw
that in Yokohama there were people sleeping behind server racks.  If
something goes wrong, heaven help you.  You now have 1500 IETF-ers that
are VERY vocal willing to tell you what you did wrong and exactly how
stupid you are.  If you are nice, they MIGHT tell you how to fix it, or
they may have to run off to a meeting (which is after all what they are
there for!).  Note: There are IETFers that are patient, polite and always
willing to help.  You know who you are and behalf of IETFers, NOCers, and
anyone who has ever used an internet protocol, thanks.

Essentially, it is a power company situation.  If you do great, no one
says anything.  If you don't do so hot, sell your stock, it's going to get
ugly.

So we've gone from people sleeping next to racks, and preparing for a year
ahead of time to no sponsors and a volunteer crew that is already taking 1
week off to go to a meeting.  Plus 1 week off to assemble the network. 
Plus 1 week off to coordinate the hardware.  What do they get?  They hope
to stay beneath the heavy and quick hand of that one vocal IPv6'er that
can't ping McMurdo Station over their Sat phone IPv4 VPN on their Timex
Sinclair.

This is not an impossible situation.  However, like many of the other
notes on this topic have stated, you need consistancy, and persistance of
knowledge.  The terminal room document is dead.  There needs to be a new
one.  But after a crew gets done doing a terminal room (donating their 3
weeks and having 6 months sucked from their soul) they probably have more
important things to do like IDs and RFCs that pertain to protocols on the
Internet, which is what they were all there to do in the first place, or
go back to their place of employment and do what they get paid to do.

Full disclosure:  I wasn't in Minneapolis this time.  I don't currently
work for the secretariat.  I'm not getting paid by anybody related to
IETF, I'm just here because I think the IETF is a cool group, and I think
I might be able to help them pursue their goals.

Cheers, good luck, and I cheerfully step back into 3rd party observer mode
until called upon.

--Brett M. Thorson

_______________________________________________

Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf

[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]