Conference 7.286::digital

    Conference JETSAM::EASYNET is probably a good place to ask.
    

    European EASYnet Manager, I believe, is Giancarlo Duella.
    
    Jim

Re 3109.0

Laurie, go to it and good luck in doing something about the network speed. It
certainly has been a bit "spotty" to say the least.  At times, lighning fast,
and at others like "snail mail".

Note well, that this is when there are big moves to make we finance types more
and more dependant on the network for closing the books and producing reports.
As ROSS/GL replaces GEARS and even CC reports are pulled across from other
countries.  

But then you know all this, don't you Laurie. Do you reckon the people making
the decisions know?

Shalom from sunny Israel
Baruch

	Indeed the European transmission network is in a period
	of changeover to "NND" (New Network for Digital) which
	is a star topology centred on Reading, UK, rather than the
	large mesh previously in place.

	This new tranmission contract (with Cable & Wireless) has
	been struck to save multiple millions $$ of telecom expense.

	In parallel with the mux'ing switchover, the DTN and the Easynet
	are in the process of being re-routed.  You will notice
	volatile behaviour until all cutovers are effected.


	Sitting in Valbonne, I am seeing it probably worse than
	you, as this was one of the first sites to cutover and we 
	are probably picking up a lot of through-routing which will
	eventually go direct to Reading.

    RE: last few.
    
    I checked the EASYNET conference, and found nothing meaningful.
    
    Baruch, how true that is. More and more we are centralising, and more
    and more we are being forced to rely on the network for our day-to-day
    running of the business. In my view, this is no bad thing, I've long
    believed our network to be one of our greatest assets. However,
    considering that our people were once our greatest asset, what's to say
    they're not prepared to strip the network down? Given the rise of
    client/server technology, some of which is even beginning to filter
    through to our internal systems, we *MUST HAVE* a fast and reliable
    network. Of course, all this applies without even mentioning WWW...
    
    RE: .4
    
    Many thanks for that information. I appreciate knowing it, and I'll see
    to it that those of us attempting to support the business understand
    why the net is so flakey recently. Yesterday, for instance, I was
    really struggling to reach Germany, and once I had, it was deadly
    slow. I'm talking of something like 300baud speed. I set host to
    Reading UK, and thence to Gernmany, and it was really, really fast. I
    was puzzled about that at the time, given my then understanding of the
    physical routing involved; your note has gone a long way to explain
    this. I, for one, would certainly have appreciated some forewarning of
    this change, and so too, I'm sure, would the customers I've been umming
    and erring to these past weeks.
    
    Again, many thanks. Could you possibly point me in the direction of
    some more information on this? Is it discussed in a conference
    somewhere?
    
    Cheers, Laurie.

	For info, contact your local site Easynet management.  Take
	up issues like performance and lack of notification with them.
	I've done so with mine.  Actually, to be fair to the Easynet guys
	and gals, this is a transmission issue - so find your local
	telecom boffin and talk about TDM path splitting :-).  If you want
	to do a thorough job, grab the DTN person as well and ask him
	to give you US secondary dial-tone :-) as compensation.

	On a brighter note, I understand that the performance impacts are
	temporary, resulting in a large concentration on the
	VBO spoke from older European mesh sites whilst other European 
	spokes at those sites are being commissioned.  NND is supposed to 
	have zero impact on current service levels once fully 
	operational.

	On an even brighter note - wrt the Easynet problems; I don't 
	believe that any other corporation even a tenth of our size
	could have achieved such a massive changeover whilst maintaining
	end-to-end network services.  Granted, it's slower, but it
	still works.  Think about it - complete swap out of physical
	infrastructure across 20-some countries involving God knows
	how many PTT's and local subcontractors, and a guy in Brussels
	is told in near-real-time by a guy (user) in the south of 
	France why his routing through UK to Bavaria is a bit slower.

	Let's sell it!

The issue raised in .0 might have nothing to do with the NND implementation.
Indeed, we have noticed since a few months an incredible increase in traffic on
EASYnet. For example, the traffic between Europe and the USA has DOUBLED 
during the last 6 months. We are looking into the causes of such a tremendous 
traffic growth but if someone has any idea, please let me know.

We all know that client/server, business centralizing their operations to 
reduce costs, home working etc require more and more networking resources. 
The funny thing is that when such strategies are proposed, very few, if none,
include the impact on telecom costs (and believe me in some cases this 
impact is pretty high). The issue is that the telecom function has received 
very stringent expense reduction goals. Thus, the telecom function is caught
between the need to reduce drastically its expenses and the need to increase
the level of services to suit the new business environment. Therefore, if
you need a better service, please request your business management to 
provide us with a solid business case so that it can be used to design a 
solution that will satisfy your needs at the best possible cost.

As mentioned in earlier replies, a new European transmission network topology 
is currently being implemented going from a mesh network to a star shaped 
network. This of course is impacting the EASYnet and DTN topologies. The 
reasons for making this change are the following:

o Reduce telecom expenditures.

o Increase manageability.

Both the above goals will be achieved by:

o Making use of the newly available public switched services (in our case ISDN)
  in order to reduce costs. ISDN, instead of leased lines, will be used to 
  provide backup services on EASYnet.

o Migrate to a star shaped topology in order to be able to manage efficiently
  the network from a remote operations center (in our case the ENOC in 
  Valbonne) and be able to adapt more quickly to the new and constantly
  changing business environment.

The NND (New Network Design), as we call it, is currently being implemented and
should be completed by early June. As .6 mentioned, this is a major project and
we tried to keep inconveniences to a minimum. Thus bear with us if you notice
some network strange behavior during the next couple of weeks. Thanks for
your understanding.

Giancarlo DUELLA
European Telecom
Data Services Manager.

    RE: .6 Let's sell it!
    
    Amen to that. Let's see marketing parade in front of the punters.
    
    RE: .7
    
    Well, there has been a huge increase in the number of people using
    Internet of late. Even a few months ago, it was only really the
    cogniscenti that knew how to do it. Couple that with the large number
    of DECcies who've left and are now communicating in and out of DEC via
    Internet, and you can account for some of it. Finally, add in WWW and
    I suspect you'll be able to explain a lot of the increase in traffic.
    These last two things have exploded in the last few months.
    
    I'll pass all the comments on to my management, and await their
    comments. Thank you one and all.
    
    Cheers, Laurie.

    <<< Note 3109.7 by GVPROD::GVA02::DUELLA "G.Duella - Eur IT Ops & Mgmt" >>>
                                  -< NND or not NND >-
    
    >>> traffic growth but if someone has any idea, please let me know.
    
    
    	Another probable cause is that Product management for some
    products, eg. InfoServers, has disappeared from Europe, with the result
    that questions have to go to the 'States.
    
    				Malcolm.

    And Laurie, if you were serious about wanting to read up on network
    plans, take a look at the attached .......
    
    /Chris.
 

                  I N T E R O F F I C E   M E M O R A N D U M

                                        Date:     27-May-1994 08:27pm CET
                                        From:     Readers Choice
                                                  CHOICE.READERS AT A1 at SALES at MRO
                                        Dept:     
                                        Tel No:   

TO: See Below

Subject: #7638-NetWise                                                          1

From  DAVE SCIUTO, LKG1-3/L12, 226-7782

You are invited to subscribe to the "NetWise" on Readers Choice.  If you 
would like to receive this newsletter it is necessary for you to subscribe 
by accessing your profile in VTX using the keyword PROFILE.  After 
verifying/selecting your organization/job function, At the MAIN MENU, 
choose "Select Topics of Interest", choose "Newsletters, Publications & 
Tools", choose Worldwide - Electronic, Scan the list alphabetically for 
NetWise, Select the number associated with it.

   			Are You Netwise?

  As a Digital employee, you use the data, voice, and perhaps video 
  network daily.  The network is an integral part of your work, aiding 
  you in communicating with the world.  Yet, how well do you know the 
  network you use each day?  Sure, you know how to send mail, use the 
  phone, or hold a meeting with other team members in other parts of the
  world, but, are you using the network efficiently?  Are you helping 
  Digital cut the high cost of telecommunications?  How well do you 
  really know the network?  Do you think that Digital's network is free?  
  If you do, you're not NetWise.

  What is NetWise?

  If you've considered ways to use the network more cost effectively, you're on 
  your way to becoming NetWise. NetWise is an ongoing internal awareness 
  campaign aimed at providing education and awareness for Digital's employees.  
  Sponsored by Digital Telecommunications (DT), the NetWise goal is to help 
  decrease telecommunications costs by creating a more responsible user culture 
  within Digital. The most common misconception NetWise is aimed at changing is 
  the idea that the network is free.
  
  Where is NetWise?
  
  NetWise information is available in a variety of media throughout 
  Digital.  Here are the places you'll find NetWise information:
  
  	o  Digital Today -- To date, several NetWise articles 	   	   
  	   containing cost-savings tips on the data and voice 
  	   networks have appeared in digital today. 
  
  	o  VTX -- Type VTX BCP at the DCL $ prompt. Choose Item 	   
  	   #17, NetWise, from the BCP menu.
   
  
  	o  NetWise information will soon be available on the 
  	   Integrated Repository (VTX IR). 
   
  	o  DVN -- NetWise tips are displayed on the screen before
  	   and after DVN broadcasts.
  
  	o  U.S. Payroll Check Stubs -- Information and tips can be
  	   found throughout the year on US check stubs.
  
  	o  Bulletin Boards -- NetWise brochures, article reprints
  	   and posters are displayed at various sites throughout
  	   the corporation.
  
  	o  NetWise Online -- There are two repository sites from 	   
  	   which NetWise information, such as brochures and posters can
             be copied.  
  	
  	   	In the US:  CTHQ::NETWISE$DIR:
  	   	In Europe:  GVPROD::IT$DOC:[000000.NETWISE$DIR]
  
  	o LiveWire -- NetWise information is occasionally available 
  	  on LiveWire.
  
    How Can I Help?
  
  	For more information about NetWise -- Contact Judi Harris 	  
         	(CTHQ::HARRIS) or Dave Sciuto (CTHQ::SCIUTO).
  	   
Distribution:
This message was delivered to you utilizing the Readers Choice delivery 
services.  You received this message because you are part of the an APA, 
European or U.S. Readers Choice subscriber.  If you have questions 
regarding this message, please contact the author of the memo.
    
Distribution: (removed)

    	It used to be One Message, One Operating system, One set of layered
    product kits.
    
    	Now, in many cases we have kits for VMS/VAX, VMS/AXP, OSF1, NT.
    Furthermore, because of staff cuts and in some cases cuts in hardware
    the voluntary network of sub distribution points has partly broken
    down, and that is going to get worse. There are more kits that people
    want, and it is becoming more likely that they can only get them
    quickly over the WAN rather than the LAN.

  Why is a star network more efficient than a distributed one? Does a "star
network centered in Reading, UK" mean that most communications between two
points on the Continent will cross the Channel twice? How can that be more
efficient than routing over the much shorter direct route? Or have I misunder-
stood what "star network" means in this context? I have some knowledge of
networks, but I'm not a guru.
  One thing I learned in the networks course I took 5 years ago (which may
already be out of date) is that a disadvantage of star networks is that if the
center is knocked out (for example, by a major storm), the entire network
shuts down. Is this the case here? 
  Also, if the network is centered in Reading, why will the network mgmt center
be in Valbonne? Wouldn't it make more sense to put the mgmt center at the
physical center? 
  This reminds me of a TV news program I saw 28 years ago when the U.S. space
program mission control was moved from Florida to Texas. The reporter asked
someone from NASA if this would cause any problems and the NASA person said,
"All the data seen by Mission Control goes over wires. This just means it will
go over a longer wire". I suppose the same is true for the Digital Europe net-
work control center. 

    	Valbonne already has a network management centre providing services
    to world-wide customer networks. Digital is just another customer (as
    it should be). If Reading can get a competitive quote from elsewhere
    then they should take it.

    re:.12
>  Why is a star network more efficient than a distributed one? Does a "star
>network centered in Reading, UK" mean that most communications between two
>points on the Continent will cross the Channel twice? How can that be more
>efficient than routing over the much shorter direct route? Or have I misunder-
>stood what "star network" means in this context? I have some knowledge of
>networks, but I'm not a guru.
    
    Yes, it's much more efficient to cross the Channel twice!  The reason
    has to do with the way bandwidth is priced.  Higher volume pipes cost
    much less per bit.  Europe's a bit more complex, but in general, any
    inter-country link is serious ECUs, and while the UK isn't
    geographially closer, its competitive telecom industry makes it a
    cheaper hub than most other spots.
    
    In the US, the price of a single 1.5 Mbps channel is roughly the price
    of two 256 kbps channels (1/3 the cost-per-bps).  Europe may not be
    quite the same, but two shares of 2-Mbps pipes to Reading (provided
    competively) is typically much cheaper than the same bandwidth obtained
    on one individual-sized (say 256k) pipe (usually from monopoly PTTs).
    In other words, it is MUCH cheaper to "buy wholesale".
    
>  One thing I learned in the networks course I took 5 years ago (which may
>already be out of date) is that a disadvantage of star networks is that if the
>center is knocked out (for example, by a major storm), the entire network
>shuts down. Is this the case here? 
    
    The Star Centre in Reading is fairly well backed up internally.  Also,
    the network in general is backed up by dial-up ISDN channels.  Dialing
    up 64 kbps channels in Europe is quick & easy, though the cost per
    minute is such that you mostly do it for backup.

>  Why is a star network more efficient than a distributed one? 

One of the issues with the previous mesh topology was that routing updates
where taking ages to propagate, because of the complexity of the mesh. So,
whenever a link was down (and unfortunately that happens from time to time),
yfinding the optimal path through the network during the transition was 
quite tough...

Also, with the previous topology we could afford smaller pipes between the
sites... With the star topology we can start putting fat pipes in place
and hence you get a much better performance (even if the links are totally
saturated as it is usually the case).

>the entire network shuts down. Is this the case here? 

No, proper contingency planning has been done and tested...

>  Also, if the network is centered in Reading, why will the network mgmt center
>be in Valbonne? Wouldn't it make more sense to put the mgmt center at the
>physical center? 

Why not put it in Valbonne ? The only choices we had were Newbury or Valbonne.
Both of them are remote. Both centers were asked to give a quote for having
them managing the switching point (since that's the name of the 'center'),
and Valbonne won. Also Valbonne was already managing all the other European
Telecom services (MTS, ELF, DNS, EASYnet routing, ...).

Once all the equipments are in place, you don't need to be physically in
the room or even near to it to manage it. If there is a hardware problem,
MCS will come and repair anyway, and not the network folks.

    Anyway, Newbury is rumoured to be for the chop soon...
    
    Laurie.

re.16:

>    Anyway, Newbury is rumoured to be for the chop soon...

I understand that, should such a decision be taken, it will be quite a task
to move all the managed networks out of there to somewhere else. The 
expense may exceed any savings.

Dave.

    Since when did common sense enter into it Dave? I heard that there are
    18mths left on the lease, and the "withdrawal" will be phased over that
    period.
    
    Laurie.

    I have been appraised of some documents available on the Net which go
    some way to explaining what is being done and why. For interested
    parties they are the following:
    
    SHIRE::ET$DNAS$DOC:S33.PS
    SHIRE::ET$DNAS$DOC:P42.PS
    
    Respectively, they are the European EASYnet Design Strategy (Oct '93)
    and the New Network Design (NND) European EASYnet Re-Design Detail
    Design Plan (Mar '94).
    
    I find it interesting to reflect that the first I heard about these
    documents, covering as they do a change that affects the European
    business in every possible way, was today... But then, I'm just a
    scumbag.
    
    Cheers, Laurie.