Re: WEATHER program, GDBM and LDM

On Thu, 26 Jul 2001, Christian Page wrote:

> One more thing about the config, I am running IRIX 6.5 on an SGI.

I don't have any experience w/ SGI's (running Solaris 7 on SPARC).  I
doubt that the Unix File System on both systems differ much at all.
Perhaps someone who has experience w/ both will comment, but it doesn't
seem like an OS related problem at first glance.

> > > The weather program is showing some inconsistency. By example, if I
> > > ask for the latest report for a given station, I get the latest report
> > > most of time, but not always. Sometimes I get the previous one, or
> > > none at all. My pqsurf.conf is configured as:
> >
> > What do you have in your 'weather.init' for 'default_time'?  The
> > distributed file does not set this and the internal default is the current
> > hour.  I changed ours to 'default_time = l', which means latest, so even
> > if the obs for the current hour hasn't arrived yet, you will still get the
> > latest obs.
> >
> I have: default_time = l12

According to the docs, this means you should get the last 12 obs.  So how
are you asking for the 'latest report for a given station'? By using the
'h1.d1-h2.d2' specifier?  Since I have 'l' as the default, I only use that
specifier when I want to look at reports earlier than the current one.
But it should work either way.

> > I don't think so, because I can ask for the latest METAR for a city at
> > around 50 minutes past the hour, get the previous hour's obs.  Then
> > without exiting the program, come back in about 7 minutes or so, and type
> > in the name of the station again (while still in the METAR> command), and
> > get the current hour's obs.  The weather program seems to access the
> > latest obs in the file with no problem and updates seem to available
> > almost as soon as they are ingested by ldm.
> >
>
> If I launch weather, ask for, let's say, /METAR then cyul, then do
> that command again and again (a few times), sometimes I get no obs, or
> sometimes I get the latest.
>
> Example of a weird output:
>
> WEATHER _V4.10_ 06-Jul-01
>
> WEATHER> /METAR
>
> METAR> cyul
> CYUL 261100Z 01011KT 15SM FEW120 FEW250 14/05 A3008 RMK AC1CI2 SLP185 =
>
> METAR> cyhu
> CYUL 261100Z 01011KT 15SM FEW120 FEW250 14/05 A3008 RMK AC1CI2 SLP185 =
>
> METAR> cyhu 2
> CYHU 260900Z AUTO 04004KT 9SM CLR 12/07 A3004 RMK SLP174 =
> CYHU 261000Z 01006KT 15SM FEW100 BKN220 10/07 A3005 RMK AC1CI2 SLP177 =
> CYHU 261100Z 05003KT 15SM FEW100 SCT220 13/09 A3007 RMK AC1CI2 SLP184 =
>
> METAR> cyhu
> CYHU 261100Z 05003KT 15SM FEW100 SCT220 13/09 A3007 RMK AC1CI2 SLP184 =
>
> All of these commands were executed one after another without a delay.
>
> Notice the second invocation where I ask for cyhu and I get cyul. If I
> ask cyhu 2 I get it correctly then if I ask again for cyhu it works
> then...

Strange.  Here's what I get:

[51] blizzard% wx
Weather Version 4.9.1 22-Sep-98

WEATHER> /metar

METAR> cyul
CYUL 261100Z 01011KT 15SM FEW120 FEW250 14/05 A3008 RMK AC1CI2SLP185=
CYUL 261100Z 01011KT 15SM FEW120 FEW250 14/05 A3008 RMK AC1CI2 SLP185=

METAR> cyhu
CYHU 261100Z 05003KT 15SM FEW100 SCT220 13/09 A3007 RMK AC1CI2SLP184=

METAR> cyhu 2
CYHU 260900Z AUTO 04004KT 9SM CLR 12/07 A3004 RMK SLP174=
CYHU 261000Z 01006KT 15SM FEW100 BKN220 10/07 A3005 RMK AC1CI2SLP177=
CYHU 261100Z 05003KT 15SM FEW100 SCT220 13/09 A3007 RMK AC1CI2SLP184=

METAR>  cyhu
CYHU 261100Z 05003KT 15SM FEW100 SCT220 13/09 A3007 RMK AC1CI2SLP184=


I didn't pick up the 'cyul' after asking for 'cyhu'.  I'm trying to
recall: possibly I have seen something like this before, but not
consistently enough for it to rise to the status of a 'problem', but I'm
not sure, it's been a while.

Sorry I can't be of more help.

Tom
-----------------------------------------------------------------------------
Tom McDermott                           Email: tmcderm@xxxxxxxxxxxxxxxxx
Systems Administrator                   Phone: (716) 395-5718
Earth Sciences Dept.                    Fax: (716) 395-2416
SUNY College at Brockport