[noaaport] 20100516: upcoming test of AWIPS NCF backup servers; new Unidata NOAAPort ingest package available for FTP download

  • To: noaaport@xxxxxxxxxxxxxxxx
  • Subject: [noaaport] 20100516: upcoming test of AWIPS NCF backup servers; new Unidata NOAAPort ingest package available for FTP download
  • From: Unidata User Support <support@xxxxxxxxxxxxxxxx>
  • Date: Sun, 16 May 2010 18:05:15 -0600
All,

A test of the AWIPS Network Control Facility (NCF) backup servers has
been scheduled by the NWS for 15:15 UTC this coming Wednesday, May 19.
The announcement of the test, broadcast in NOAAPort and relayed to
users subscribed to our nws-changes@xxxxxxxxxxxxxxxx email list, is
included at the end of this email for your review.

Those of you subscribed to this list who are operating NOAAPort ingest
systems using the Unidata 'noaaport' package will be aware that the
past several tests of the NCF backup servers have resulted in
"corruption" of some nwstg2 GRIB2 messages (typically multiple GRIB2
messages were sent as a single product).  Some of the products have
been seen to cause 'noaaport's processing routine 'readnoaaport' to
exit with error, and this, in turn has led to the LDM running on the
ingest machine exiting.  The NOAAPort ingest machines run by the
UPC have been instrumented to monitor data ingest, and restart the
LDM upon unintended exit.

Code modifications and GRIB2 table updates made by Steve Chiswell
(DOE/Savannah River National Laboratory) have been included in a new,
** TEST ** release of our 'noaaport' ingest package, noaaport-1.4.0.tar.gz.
This release is available by FTP as follows:

machine:   ftp.unidata.ucar.edu
user:      anonymous
pass:      <your full email address>
directory: pub/noaaport
file:      noaaport-1.4.0.tar.gz

Please be aware of the following changes to the code and their possible
impact on you and/or your users:

1) product bulletin sizes are compared to decoded GRIB lengths to 
   trap and prevent overruns (the cause of problems see in the current
   release)

2) SDUS8 products are being moved from HDS to NEXRAD3

   This allows future dual-polarization products to be included in the
   NEXRAD3 datastream, and it will cause the SDUS DPA products to move
   from the HDS to NEXRAD3 datafeeds.

   Implication:

   You, and/or users who feed from you may have LDM pattern-action file
   actions that process DPA products from HDS.  After the new code is
   up and running, sites using the DPA products will need to modify
   their setups to process the DPA products from the NEXRAD3 feed.

3) the g2clib library bundled with 'noaaport' was updated to version
   1.2.0 to add the WAFS (thinned grids) coming in the near future

4) GRIB2 decoding tables were updated for new levels and parameters

5) an argument was added to 'dvbs_multicast' to specify the multicast
   interface (in case a static route is needed instead of the default
   route)

6) readnoaaport-1.4, like its predecessors, uses LDM configuration
   files for its build.  readnoaaport-1.4 has been tested with
   LDM-6.8.x only.  It is not expected to work with very old
   LDM installations, and it is known to NOT build on systems
   where the new test LDM-6.9.x is installed.

   The build setup for noaaport-1.4 will need to be updated to
   work with LDM-6.9.x.  Until a new release of noaaport is made,
   you are advised to continue using LDM-6.8.1.

NB:

Because of 2) above, we are requesting that sites running NOAAport
ingest systems whose output is fed into the IDD wait until the late
afternoon or evening of Tuesday, May 18 to transition to the new code.
On Monday, May 14 an email will be sent to the ldm-users@xxxxxxxxxxxxxxxx
email list notifying IDD participants of the impending change and
advising them of the action(s) that they will need to take to minimize
impacts to their data processing.

Please relay this message to folks you know are running the Unidata
NOAAPort ingest package but not subscribed to this list.

Please direct questions/comments to support-noaaport@xxxxxxxxxxxxxxxx.

Cheers,

Tom Yoksas
Unidata User Support


The following is the NWS announcement of the NCF backup test:

196 
NOUS72 KNCF 141548
ADMNCF
 .
PLEASE PASS TO THE MIC/HIC, SOO, ESA, ITO AND AFP: 
. 
THERE WILL BE AN OPERATIONAL TEST OF THE AWIPS 
NETWORK CONTROL FACILITY (NCF) BACKUP SERVERS.  
NCF SERVER OPERATIONS WILL BE SWITCHED FROM THE 
PRIMARY AWIPS NCF (ANCF) IN SILVER SPRING, MD TO THE 
BACKUP NCF (BNCF) AT FAIRMONT, WV.   THE SWITCH IS 
SCHEDULED FOR WEDNESDAY, 19 MAY AT 1515 UTC.  
OPERATIONS WILL BE SWITCHED BACK TO THE PRIMARY 
NCF ON THURSDAY, 20 MAY AT 1515 UTC.  THE SWITCH 
WILL NOT BE DONE ON A CRITICAL WEATHER DAY, OR 
IF THERE ARE OTHER SIGNIFICANT OPERATIONAL ISSUES 
WARRANTING POSTPONEMENT.  
. 
ALL ANCF MESSAGE DISTRIBUTION WILL BE REPOINTED 
TO THE BNCF.   THE SWITCH TO THE BNCF WILL BEGIN 
AT 1515 UTC AND WILL TAKE ABOUT 20 MINUTES.  DATA 
WILL QUEUE WHILE THE SWITCH IS IN PROGRESS AND 
WILL BE SENT ONCE THE SWITCH IS COMPLETE.  THIS 
WILL AFFECT DATA ORIGINATING AT NWS FIELD OFFICES 
AND NATIONAL CENTERS ON THE AWIPS NETWORK 
INCLUDING RADAR DATA. 
. 
REPOINTING OF OPERATIONS BACK TO THE ANCF WILL 
ALSO TAKE ABOUT 20 MINUTES DURING WHICH DATA 
WILL BE QUEUED.   QUEUED DATA WILL BE SENT WHEN 
THE REPOINTING IS COMPLETE. 
 .
NCF

--
********************************************************************** <
Unidata User Support                              UCAR Unidata Program <
(303)497-8643                                            P.O. Box 3000 <
support@xxxxxxxxxxxxxxxx                             Boulder, CO 80307 <
---------------------------------------------------------------------- <
Unidata Web Support               http://www.unidata.ucar.edu/support  <
********************************************************************** <



  • 2010 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the noaaport archives: