[IGSMAIL-012] DATA flow start 15 JUNE

Wed Jun 10 08:56:37 PDT 1992

***********************************************************************
IGS Electronic Mail       10-JUN-1992 08:56:37       Message Number 12
***********************************************************************

==============================================================================
                   THIS FILE CONTAINS 2 MESSAGES
==============================================================================

>From:    Peter Morgan
Subject: DATA flow start 15 JUNE
         -----------------------

Dear Ericos
Thankyou for your note on the data flow.
I am feeling much better at the moment. The intense pain has subsided
and my general well being is improving rapidly.

I agree with you on starting the data flow as early as possible.
As you state Carey is due back this week. I will be trying to call early
today. I spoke to John Bosworth on Friday and he acknowledges the
importance of getting this to the top of Carey's priority list.

There are two items that need to be straightened out for data flow.

1. What is the EXACT path of entry of the following stations into A
NETWORK DATA CENTER. At the moment this is SIO/SCRIPPS, CDDIS and
STATENS KARTENS.

       CANBERRA (TIDBINBILLA)       JPL
       Yaragadee                    JPL
       Madrid                       JPL
       Goldstone                    JPL
       Tahiti     
       Hartebeestoeck
       Santiago                     JPL
       Usuda                        JPL
       Fairbanks                    JPL
       McMurdo                      JPL
       Pinon Flat                   SIO
       Los Palmas
       Matera
       Kootwijk
       Ny Alesun
       Tromso
       Onsala
       Metsahovi
       Wettzell
       Herstmonceux
       Hobart
       Wellington
       Townsville
       Koke Park
       Yellowknife
       Penticton
       Algonquin
       Richmond
       Fortaleza
       Taipei (Taiwan/Tai Shi)
       Port Harcourt
       Tsukuba
       Orroral 

This list reflects the Circular 2 list although we dont expect all sites
to be equipped with precision P-code receivers. At this point in time,
knowing that our data network can handle the load I am reluctant not to
collect this full set leaving the analysis centers freedom to choose.

I point out that Orroral is now an Ashtec dual P-code. I will initiate
discussions to see if this is indeed the only place that it can be run
for the campaign. From a geometry point of view Cocos Is is the best
location. We OZ folk will run that site for Epoch'92 but cannot afford
operational costs for the main campaign. I presume that Darwin would be
acceptable for many folk. 

I think that the correct way to fill out the above is for Carey, in
Miranda's absence to fill in what she got from NGS. I presume that
Tahati and Hartebeestoeck will come from JPL. I would appreciate IGN
explicitly stating that this is the start-up path.

2  Compression routines. I think that the easiest way is for Carey to
send all of us a copy of a file that she has compressed. We should all
be able to say that we can read it. As I understand things now she got
her routines from Werner. Werner, SIO, Myself and JPL can all transfer
files without problems. Its just a case of getting the switches right. I
dont think that we should interfer with Carey and Dave Starr doing
uncompressed transfers since their bandwidth allows this. However I
think that our problems of congestion will be lower if we do adopt this
strategy and do it as close to the siurce as possible.

I will call Carey tonight about 11pm local time here. That is 1300UT.

The University of Canberra, in conjunction with the regional data center
at AUSLIG will collect a complete copy of CORE data and maintain a
complete copy of ALL regional data. This will reduce the load on the
PACIFIC link. This has recently been upgraded to 512 kbits per second.
In this upgrade there was the express desire to limit FTP traffic as
much as possible. AARNET has placed a small file server on the AARNET to
reduce multiple US accesses. We are being encouraged to do the same. For
IGS this is similar to our desire to have a network data center in
Europe.

Finally there is a lot of development going on on the network about
mailers. Thankyou for your comments on the LISTSERVER approach which is
used by UNB for their space forum. I hope that I can sumarize some of
this shortly.

regards  peter

==============================================================================

>From:    Loic Daniel
Subject: IGN network data center
         -----------------------

Here is the current status of the data center at IGN.

First of all, we expect to get our connection to the Internet for the 15th 
of june. As soon as this connection is available, I intend to do some data
transfer tests with some european centers (CNES, IFAG, SK an CODE) and with
the american network data centers: CDDIS and SIO.

The data rate of the connection is 64 Kbps, this is the first parameter I want
to check, then I also need to verify the procedures we have developped to
manage the data flow in a fully automatic way.

At IGN, during the IGS campaign, we will first GET the data from IFAG and
CNES, then GET the complementary data of the core network from CDDIS and SIO.
I don't know if CODE intends wetherto receive the data from IGN or to get it
from IGN, can Werner GURTNER answer this question ?

The on-line storage capacity is 1 GB, so depending on the data compression rate
we should be able to keep between 40 and 80 days on-line. We will archive the
data on optical disks and DAT. The operating system is VMS. The data is stored
in DAY/STATION directories, the root being IGS$DATA:[CORE]. We will provide
access to the data by FTP under userid IGS and password REPUBLIQUE. The IP 
address will be 192.33.147.225.

To summarize the status of IGN, we are  ready to operate as soon as we are
connected to the Internet. I think that one week is a very short delay to 
get fully operational on the IGS network, and that will be the maximum 
delay for us considering the date of the 15 june for the connection. So we
might have to consider a temporary configuration without IGN as network 
data center. In any case this should not last more than 15 days after the
beginning of the campaign. I will send more information on this to the mail 
box next week. Can George PREISS from SK confirm their ability to operate
as a network data center during this period ?

My last point is that I need more information about the other IGS centers
(for some of them I don't have any), particularly those with wich we will
operate directly. I've heard of a directory of all the centers 
participating to the IGS campaign but I don't know if it is achieved. 
Anyway, this would be a very useful document for all of us. Can CDDIS, SIO
and CNES mail me their IP address, FTP account, computer directory names. 
Do they provide access to the data 24 hours a day ?

Regards

Loic Daniel