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

Re: notifyme - first look

Harry Edmon wrote:
> 
> I ran notifyme on thelma for approx 24 hours.  When I started to look
> at the output, I was surprised to see that notifyme could not keep up
> with the feed - it eventually got 4 hours behind, so I may have missed
> some products.!  Now that I think about it, having one notifyme for the
> whole thelma feed was a mistake  - there are too many RPC calls for one
> process to keep up with.  And it is hard to seperate out the different
> feeds.  I think I run several notifymes with the same feed types and
> patterns as my thelma ldm server feeds.
> 
> P.S. How many hours of products are currently kept in thelma's queue?
> --
> Dr. Harry Edmon                 E-MAIL: address@hidden
> 206-543-0547                            address@hidden
> Dept of Atmospheric Sciences    FAX:    206-543-0308
> University of Washington, Box 351640, Seattle, WA 98195-1640

Hi Harry,

First, FYI, we lost connectivity with thelma for at least 15 minutes
this morning, roughly between 15:25 and 15:40Z.  The networking people
are upgrading our routers and had an oops.

I have not yet tried to run notifyme on a full feed for that long.  I'll
take your advice and run several.  Currently I see on thelma 7
notifyme's to sunny.

Yesterday I wrote Python code to run notifyme, write the output to a
file, and close and start a new log file based on a SIGALRM.  The
SIGALRM comes from a script run by cron, so it could occur at any time. 
I'm nearly finished with a script to grab the signatures from one file
and ensure that each signature appears once and only once in any other
notifyme outputs.  I hope to finish all this today.  Do you have Python? 

Over the past week thelma's queue has held a minimum of 2.6 hours to a
max of 5.6 hours worth of data.

On to your next message...

Anne
-- 
***************************************************
Anne Wilson                     UCAR Unidata Program            
address@hidden                 P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************