[Go-essp-tech] wget errors

[Estanislao Gonzalez]

Hi,

If this emails went to go-essp-tech, I can't figure out why I haven't 
got them... (well, perhaps I'm on the spam list now, can't figure out 
why :-)

I'm working indeed to merge this two scripts version, I'm sadly still 
not done as my attention keeps jumping from one thing to another.
But I'd like to comment on what I am doing and where I'm heading to get 
some feedback.

I'll comment in-line:

Am 17.05.2011 15:42, schrieb stephen.pascoe at stfc.ac.uk:
> Note, Estani has been working on the script template.  I sent him our version yesterday so we can merge the two together.  He should be kept in the loop on discussions on how to improve the script.
>
> S.
>
> ---
> Stephen Pascoe  +44 (0)1235 445980
> Centre of Environmental Data Archival
> STFC Rutherford Appleton Laboratory, Harwell Oxford, Didcot OX11 0QX, UK
>
>
> -----Original Message-----
> From: go-essp-tech-bounces at ucar.edu [mailto:go-essp-tech-bounces at ucar.edu] On Behalf Of philip.kershaw at stfc.ac.uk
> Sent: 17 May 2011 14:38
> To: Lawrence, Bryan (STFC,RAL,RALSP)
> Cc: go-essp-tech at ucar.edu
> Subject: Re: [Go-essp-tech] wget errors
>
> Hi Bryan,
>
>> I just me a bloke here at ICTP who is struggling to download data, he's
>> sending me details, but I suspect it's due to certificate timeouts.
>>
>> 1) does the wget script give an error message which is comprehensible to
>> "joe-average-scientist", and
> I think that would get messy as you'd need to heavily customise the
> script.  I think the best approach is to make it clear to users (through
> the help documentation and UIs) that in the event of problems they should
> send a query to the support list with the full output from the script.
This is "half" done, and I think we can expand it a little further. At 
the moment (pre-merging with Stephen's one) the script does test if the 
user has an expired certificate (in this case it aborts everything with 
a proper message) or if it will expire soon (in which case it just warns 
the user, but this might get overlook, should we expect some action from 
part of the user no continue?)
We could trap the script's exit and recheck the certificate there, so 
the last line the user will have will be a comment regarding this, if 
that's the case. Is this ok?

>
>> 2) how does one get a new certificate ... if one already has one and it's
>> expired ... (it seems like he is not prompted to do the java download
>> the day after he has one working).
> You can fire up the Java MyProxyLogon at any time and get a fresh
> certificate.  On Linux and Mac desktops I've tried, you can double click
> on the downloaded program icon to restart it.
>
> We've debated whether to include the MyProxyLogon call actually in the
> script itself if a certificate has expired but there's a balance to be had
> between bloating the script with too much additional logic.
Well the logic will be to download the Jar file (I have already a 
version that does not require any gui) and start it directly. This 
sounds easy (it is actually) but the problem comes from the code to make 
it robust (what happens if no valid java version is installed, and, and, 
and....)
So yes, this will slowly, but continuously, increase the script complexity.

>> This guy is also one who is completely puzzled by the difference between
>> PCMDI and BADC wget script behaviours.
> Bob has said he's committed to switch over the interface so that we're
> consistent with one another.
>
> Cheers,
> Phil
>
Ouch, well we don't have any data, but our script is... again... 
different. Hmmm....

The script is intended as a simple "Work-around" as there are a couple 
of issues preventing it from being used as a real shell-command, like 
it's impossible to download it remotely via wget, so if you are running 
it somewhere else, you have to copy&paste it's contains (or the whole 
file) to the remote machine.

In any case I expect people to use DML if they are working with anything 
but the shell. The script was always a "Plan B" idea.

If we are going to proceed further in this direction I suggest we host 
the script centrally (say esgf.git?), version it and display the current 
version to the user always (as well as the esg-support email address and 
encourage the user to mentioned it when sending an email in case of 
trouble).

Does this sound ok?

Thanks,
Estani
>> --
>> Bryan Lawrence
>> Director of Environmental Archival and Associated Research
>> (NCAS/British Atmospheric Data Centre and NCEO/NERC NEODC)
>> STFC, Rutherford Appleton Laboratory
>> Phone +44 1235 445012; Fax ... 5848;
>> Web: home.badc.rl.ac.uk/lawrence
>> -- 
>> Scanned by iCritical.


-- 
Estanislao Gonzalez

Max-Planck-Institut für Meteorologie (MPI-M)
Deutsches Klimarechenzentrum (DKRZ) - German Climate Computing Centre
Room 108 - Bundesstrasse 45a, D-20146 Hamburg, Germany

Phone:   +49 (40) 46 00 94-126
E-Mail:  gonzalez at dkrz.de