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

Re: 881109: Re: 19991104: RE: GEM Model Data (fwd) (fwd)

> ---------- Forwarded Message ----------
> Date: Tue, Nov 9, 1999 3:07 PM -0700
> From: Linda Miller <address@hidden>
> To: Ekaterina Radeva <address@hidden>
> Subject: 881109: Re: 19991104: RE: GEM Model Data (fwd)
> 
> Hi Ekaterina,  
> 
> It is good to hear from you.  See below.
> 
> 
> --On Tue, Nov 9, 1999 4:34 PM -0500 Ekaterina Radeva
> <address@hidden> wrote:
> 
> > Dear Ms. Miller,
> > 
> > I am writing you in connection of the installation of the LDM software at
> > the Canadian Meteorological Center. I am part of the Implementation and
> > Operational Services Division headed by Richard Hogue. Actually, I am more
> > of a meteorology than a computer support person, but with sufficient
> > background to deal with the installation/configuration issues from an user
> > point of view. 
> 
> That sounds good.
> 
> > 
> > I read with great interest the information you have posted on the UNIDATA
> > web page. It is very well written and do take you step by step through the
> > installation process. However, being a national center, we have strict
> > protocols about installing new software. The "by-the-book" way to do this
> > will be to submit a service request to our Informatics Branch where all
> > new software installation and testing is handled. This formal procedure
> > takes more time that depends on their work load at the moment. Anyway,
> > that's the only way to go, as I am not allowed to act as "root" on the
> > machine that is designated as LDM host.
> > 
> 
> Unidata has dealt with various problems with installation, and I have copied
> this message to the Support staff at Unidata, who should be able to help you
> with installation questions that you may encounter.
> 
> 
> > The primary concern of our colleagues at Informatics is the security
> > issue. They need to make sure that LDM cannot be exploited to break into
> > the operational network. We plan to install the package on a server that
> > is external to the "firewall". The "firewall" is a security system that
> > allows only unilateral connection between a machine and the operational
> > network, i.e. one can connect/transfer files from a operational machine to
> > the external server but not the other way. I guess that will be sufficient
> > from a security point of view ?

Ekaterina,

Any piece of software can be a security problem. The LDM has been designed
to be secure by using a predefined set of protocols.  These protocols
only accept a certain set of commands:

http://www.unidata.ucar.edu/packages/ldm/ws/ldm_protocols.html#s6

For someone to break into your machine they would have to exploit these
commands. So far it hasn't been done, the ldm is running on machines at
160 sites and many government type agencies. 

With that said, the machine that the LDM runs on needs to be as secure as
possible. Details at:

http://www.unidata.ucar.edu/packages/ldm/ws/ldm_protocols.html#s5
> >
>   
> I will defer this question to the Unidata support staff.
> 
> 
> > The other issue I wanted to consult you about is the content and format of
> > the data to be eventually distributed via LDM. We would like to know what
> > the other centers are distributing (variables, area of coverage, temporal
> > and spatial resolution, etc) to guide us as to what is of interest to the
> > users. I looked at the Unidata web page, but couldn't find anything
> > specific. Could you provide us with some data samples from the NCEP and/or
> > ECMWF model output products, or point us to some information about them ?
> > 
> > Also, there is the question about the data format, i.e. headers needed by
> > the LDM to identify the product, and feedtype.

At this time, you could use the EXP feedtype and then make the products ID
so the end user could key on certain variables, coverage, temporal and
spatial resolution.

Here's what the NCEP product ID's look like:

Nov 10 22:22:50 pqutil:    15024 19991110215342.023    NMC2 257
/u/ftp/gateway/ncepe/nggrib/ruc2a.991110/ruc2.T21Z.bgrbanl
!grib/fsl/105/#236/199911102100/F000/VVEL/hybrid lev 18! 000257
Nov 10 22:22:50 pqutil:    36352 19991110215748.980    NMC2 085
/u/ftp/gateway/ncepe/pggrib/ruc2a.991110/ruc2.T21Z.pgrbf01
!grib/fsl/105/#236/199911102100/F001/RH/725 mb! 000085
Nov 10 22:22:53 pqutil:    12892 19991110215342.136    NMC2 261
/u/ftp/gateway/ncepe/nggrib/ruc2a.991110/ruc2.T21Z.bgrbanl
!grib/fsl/105/#236/199911102100/F000/VVEL/hybrid lev 22! 000261
Nov 10 22:22:59 pqutil:    10758 19991110215342.220    NMC2 264
/u/ftp/gateway/ncepe/nggrib/ruc2a.991110/ruc2.T21Z.bgrbanl
!grib/fsl/105/#236/199911102100/F000/VVEL/hybrid lev 25! 000264
Nov 10 22:22:59 pqutil:    19290 19991110215342.641    NMC2 280
/u/ftp/gateway/ncepe/nggrib/ruc2a.991110/ruc2.T21Z.bgrbanl
!grib/fsl/105/#236/199911102100/F000/CLWMR/hybrid lev 1! 000280
Nov 10 22:23:00 pqutil:    34220 19991110215749.426    NMC2 108
/u/ftp/gateway/ncepe/pggrib/ruc2a.991110/ruc2.T21Z.pgrbf01
!grib/fsl/105/#236/199911102100/F001/RH/150 mb! 000108
Nov 10 22:23:00 pqutil:    10758 19991110215342.283    NMC2 267
/u/ftp/gateway/ncepe/nggrib/ruc2a.991110/ruc2.T21Z.bgrbanl
!grib/fsl/105/#236/199911102100/F000/VVEL/hybrid lev 28! 000267


Nov 10 22:23:01 pqutil:    36352 19991110215749.051    NMC2 089

This is machine time, program pqutil, size, arrival time, feedtype, seq #

/u/ftp/gateway/ncepe/pggrib/ruc2a.991110/ruc2.T21Z.pgrbf01
!grib/fsl/105/#236/199911102100/F001/RH/625 mb! 000089

Actual product ID.

Looks like directory structure, model time, model info, time, parameter RH
and level 625.

This doesn't mean you have to follow this structure, but think about how
the end users would want to receive certain products in the data stream.

If you send some information on the parts of GEM stream, we could possibly
help with creating the products ID's


Robb...
> 
> The above two paragraphs should be handled by the Unidata support staff, who
> have dealt with many of these issues.
> 
> 
> > 
> > These are a lot of questions for an introductory message, but there is a
> > great interest at CMC to get to the Unidata "highway".
> 
> Please keep me informed on your progress of making the GEM data available to
> the broader university community.  Richard and I will need to discuss the
> access and distribution of the data, after you and the Informatics Branch have
> the LDM running.
> 
> > 
> > Thank you in advance for your help.
> 
> Thanks!
> 
> Linda
> 
> > 
> > With best regards,
> > Ekaterina 
> > 
> > 
> >  
> > 
> > ______________________________________________________
> > 
> > Division de l'implementation et services operationnels
> >    Centre meteorologique canadien
> > Implementation and Operational Services Division
> >    Canadian Meteorological Centre
> > 
> > e-mail: address@hidden
> > tel: (514) 421-4646
> > ______________________________________________________
> > 
> > On Thu, 4 Nov 1999, Linda Miller wrote:
> > 
> >> To:  Richard Hogue, Canadian Meteorological Center
> >> 
> >> Hi Richard,
> >> 
> >> 
> >> --On Thu, Nov 4, 1999 11:07 AM -0500 "Hogue,Richard [CMC]"
> >> <address@hidden> wrote:
> >> 
> >> > Hi Linda,
> >> > First of all, sorry for not getting back to you sooner.  The whole Y2K
> >> > preperation area is taking a lot of everyone's energy.  We indeed have
> >> > started to look into LDM in more detail but have not got it installed in 
> >> > a
> >> > formal way on our external server.  In my Division, the follow-up work 
> >> > will
> >> > be done by Ekaterina Radeva.  If this is OK, she will be contacting you 
> >> > for
> >> > some specific technical questions and for some samples.
> >> 
> >> This sounds great.  As you can see, I have copied Unidata support on this
> >> email.  We usually ask our users to contact support
> >> (address@hidden) with questions; they forward the email to the
> >> appropriate person for response. This will be the case for this
> >> collaborative activity too.  I did copy Robb Kambic on this email because,
> >> at this time, he will probably be the technical contact who will respond to
> >> specific LDM-related questions from Ekaterina, however she may receive
> >> information from others if Robb is away.
> >> 
> >> I'll provide the technical information that you have shared to our support
> >> team, and I'm sure Ekaterina will be able to install the LDM without a lot 
> >> of
> >> trouble by following the LDM Site Manager's Guide provided at:
> >> 
> >> http://www.unidata.ucar.edu/packages/ldm/index.html
> >> 
> >> As progress is made, we'll need to coordinate the process of availabilitly 
> >> of
> >> the data to university sites.  That can happen later though.
> >> 
> >> > Please copy me as well in your replies to Ekaterina.
> >> > Thanks for you extra patience...we're going to get there.
> >> > 
> >> > Thanks for the info on the AMS annual meeting.  I will not be attending
> >> > myself but a few of my colleagues will.  I will make sure that someone 
> >> > does
> >> > go to the Unidata session.
> >> > 
> >> 
> >> We'll be looking for your colleagues at the AMS.
> >>  
> >> > Best regards,
> >> > Richard.
> >> > 
> >> 
> >> Thanks again,
> >> 
> >> Linda
> >> 
> >> 
> >> > 
> >> >   ---------- Forwarded message ----------
> >> >   Date: Thu, 07 Oct 1999 10:20:32 -0600
> >> >   From: Linda Miller <address@hidden>
> >> >   To: address@hidden
> >> >   Cc: address@hidden
> >> >   Subject: GEM Model Data
> >> > 
> >> >   Hi Richard,
> >> > 
> >> >   I haven't heard from you for some time and wondered if your folks
> >> > have tested
> >> >   the LDM for handling the GEM data.  Let me know how things are
> >> > progressing of
> >> >   if you wish to propose another option for consideration.  
> >> > 
> >> >   Are you planning to attend the AMS Annual Meeting this year?
> >> > Unidata is having
> >> >   IIPS-Unidata Special Session on Tuesday, 11 January 2000.  The
> >> > schedule 
> >> >   for presenters has been published by the AMS.  Look for Session 5:
> >> > Unidata
> >> >   Applications and Extensions.  The session has attracted many fine
> >> > presenters
> >> >   and, if you have time, you might want to catch a couple of the
> >> > presentations.
> >> >   In fact, if you're attending, perhaps we should attempt to set up a
> >> > meeting to
> >> >   discuss the GEM model distribution.
> >> > 
> >> >   Linda
> >> > 
> >> > 
> >> >   Linda Miller - address@hidden
> >> >   External Liaison, Unidata
> >> >   University Corporation for Atmospheric Research
> >> >   P.O. Box 3000
> >> >   Boulder, CO 80307-3000
> >> >   303 497-8646 fax: 303-497-8690
> >> >   URL:  http://www.unidata.ucar.edu/staff/lmiller/un.act.html
> >> 
> >> 
> >> 
> >> Linda Miller - address@hidden
> >> External Liaison, Unidata
> >> University Corporation for Atmospheric Research
> >> P.O. Box 3000
> >> Boulder, CO 80307-3000
> >> 303 497-8646 fax: 303-497-8690
> >> URL:  http://www.unidata.ucar.edu/staff/lmiller/un.act.html
> >> 
> >> 
> > 
> 
> 
> 
> Linda Miller - address@hidden
> External Liaison, Unidata
> University Corporation for Atmospheric Research
> P.O. Box 3000
> Boulder, CO 80307-3000
> 303 497-8646 fax: 303-497-8690
> URL:  http://www.unidata.ucar.edu/staff/lmiller/un.act.html
> 
> ---------- End Forwarded Message ----------
> 
> 
> 
> Linda Miller - address@hidden
> External Liaison, Unidata
> University Corporation for Atmospheric Research
> P.O. Box 3000
> Boulder, CO 80307-3000
> 303 497-8646 fax: 303-497-8690
> URL:  http://www.unidata.ucar.edu/staff/lmiller/un.act.html
> 

===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================