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

20020308: setting up request lines in ldmd.conf (cont.)

>From: "Kevin Polston" <address@hidden>
>Organization: NOAA/NWS
>Keywords: 200203050151.g251p7K03671 IDD

Kevin,

After reading through your original and then follow-up note, all I can
suggest as the cause for ingest failure of NIMAGE images is that when
you shut down the LDM the first time, you did not wait until all LDM
processes had exited before restarting it.  This is an mistake thing to
make since the ldmadmin script returns back and says that the LDM
has been stopped.  What one has to do at that point is a ps that
looks for rpc.ldmd processes:

ps -eaf | grep rpc.ldmd

If you see any of these processes, you should wait until they exit
before restarting the LDM with 'ldmadmin start'.  In general, you
should _not_ kill those rpc.ldmd processes since you face the
possibility of corrupting the LDM queue by doing so (doesn't happen
all of the time, but it does happen).

Successful resumption of ingest and filing of the NIMAGE products after
a reboot goes a long way towards confirming my suspicion on the lingering
rpc.ldmd process(es).

>I have a problem here and can't figure it out. (it is probably an
>"operator error" type of deal).  Last evening....right after 00Z I did a
>rm -R command on my sat directory   _20020306   ...actually I ran it
>from this directory here /usr1/nawips/metdat/images/sat    because my
>disk was getting filled up.

Have you not implemented the data scrouring of the image directories as
we discussed in a previous email?  Correctly setting up scouring will
prevent you from having to manually delete files.

>Well I have done this before with little or
>no problems...and then a new directory gets created when the new data
>starts flowing in. Well I have not been saving any imagery this morning
>and I can't figure out why. It looks like teh satellite data is sent
>from papagayo but I'm not storing it for some reason. I slightly altered
>my ldmd.conf file (and stopping and restarting the ldm to invoke the
>file) but it didn't seem to take affect.  I am sending you my ldmd.conf
>files and my pqact.conf file (although I never touched it) to take a
>look and see if I have doen something wrong. I am almost certain that
>they are both ok. The only question I have is if I set up the ldmd.conf
>file correctly. I think I have it formatted properly.  So perhaps you
>can shed some light on what I have done wrong or what might be the
>problem. 
>
>Also, I have not received any profiler data since 01Z. And the radar and
>surface obs are running behind again....but I figure that might be
>related somehow to my current problem. So if you have any answers let me
>know...it seems all I have are questions.  :-) 

I would not expect the surface obs to start running late after a successful
'ldmamdin stop"; wait; 'ldmadmin start'.  After you reboot, however, if
your startup procedure includes deleting and remaking the queue (the
safest course of action on a reboot) then I would expect just about
everything to be late.  The reason for this is that when you remake your
queue and then start the LDM, it will request all data for the previous
hour from its upstream host(s).  It will then have to ingest all of this
data before getting to the current stuff.

>Looking forward to tomorrow...possibility of severe around here. I hope
>things are running smoothly by that time.
>
>KP
>
>>From address@hidden Thu Mar  7 13:13:38 2002
>>Subject: Re: 20020305: setting up request lines in ldmd.conf
>
>Tom,
>
>OK, I don't know what happened but here is what I did. I restarted my
>machine....and now...all of a sudden it lookslike satellite imagery is
>coming in.  It also appears to be just the EAST-CONUS and WEST-CONUS as
>I wanted from my changes to the ldmd.conf.  Why did that happen?

See above.

>I mean,
>I rebooted the machine earlier after I had made changes,

Given this, I have no explanation for the problems.

>I did the
>start/stop of the ldm and still no luck. What the heck happened which
>got things started again?

I don't know.  At times like this I normally would request a login to
a user's machine to look at the actual setup, but since I can't even
ping your machine this is not possible.

>I am frustrated mainly because I don't
>understand what happened here. I am happy the data is flowing in again 
>:-)  but I guess I am not up to the power user level yet (or as close as
>I thought I was getting).

You may have to bite the bullet and read the LDM manual :-)

>Well....I thought I would let you know that
>things are getting back to "normal".  Hopefully that will stay the case. 

I hope so as well.

>Talk to you later,

Later...

Tom