RFC ARCHIVE DATABASE Bug List

Current Bug Fix Requests-Unassigned
Current Bug Fix Requests-Assigned
System Fixes
Bugs Fixed For Release OB-3
Bugs Fixed For Release OB-4
Bugs Fixed For Release OB-5
Bugs Fixed For Release OB-6
Bugs Fixed For Release OB-7
Enhancement Requests-Unassigned
Enhancement Requests-Assigned
 


Current Bug Fix Requests-Unassigned

Last Updated:   7/14/05
New Additions to Latest Update:  r1-58


Current Bug Fix Requests-Assigned

Last Updated:  7/14/05


RAX-shef_decode_pro – The raw shefdecoder is not doing the datalimits/locdatalimits testing that it is supposed to be doing.  When the shefdecoder processes a piece of data that should be flagged as "B" in the PECRSEP table, but instead it is being flagged as "G".
     Note a test message for the raw shefdecoder can be found in //fs/hseb/bugs/r26-/batchoutput_pairing.txt

        HSD Bug #:  r1-57  Programmer:  OHD
        Original RFC           Entry  Date       Note
        APRFC                     7/11/05


RAX-processed_stage – Data is not flowing consistently into the processed tables, in spite of what is indicated in the log.

        HSD Bug #:  r1-56  Programmer:  ba
        Original RFC           Entry  Date       Note
        APRFC                     7/11/05


RAX-get_params – When attempting to run the "get_params" program to load data into the appropriate tables in the archive database, the program terminates with an SQL error less than half way through the load.

        HSD Bug #:  r1-55  Programmer:  ba
        Original RFC           Entry  Date       Note
        NCRFC                     7/11/05


RAX-shef_decode_raw – The raw shefdecoder is not doing the datalimits/locdatalimits testing that it is supposed to be doing.  When the shefdecoder processes a piece of data that should be flagged as "B" in the PECRSEP table, but instead it is being flagged as "G".
     Note a test message for the raw shefdecoder can be found in //fs/hseb/bugs/r26-45/sheftest.txt

        HSD Bug #:  r1-54  Programmer:  OHD
        Original RFC           Entry  Date       Note
        ABRFC                     6/15/05


RAX - SHEF_DECODR_RAW, SHEF_DECODE_PRO - Dates with date-time prior to Jan 01, 1970 is not posting properly to the archvie db even though the shefdeocers posting summary information indicates that the data posted.  Some preliminary testing show the currentl decoders appear to change any year that is before 1970 to 1970 when it posts the date.  The ob1-ob3 shefdecoders posted the data just fine.
     Note a test message for the raw shefdecoder can be found in /rfc_arc/test-ajm, brii4.test or brii4stg.shef..  For the processed shefdecoder just need to alter the
     shef code to be HGIPZ.  The files brii4_elg.unl and brii4_loc.unl provide load info for ingestfilter and location tables for this one location.

        HSD Bug #:  r1-53  Programmer:  OHD
        Original RFC           Entry  Date       Note
        NCRFC, APRFC      5/26/05


shef_decode_raw – The raw decoder is doing some extra checks based on the SHEF data qualifier code which shouldn't be occurring...it's been noticed particularly with HADS data.  When the SHEF data qualifier code is certain values the shefdecoder throws the data way instead of posting it.  The raw decoder should behave in a similar manner to the IHFS shef decoder.

       ADB Bug #:  r1-51      Programmer:  OHD
        Original RFC             Entry  Date        Notes
       MBRFC                       3/22/05


shef_decode_raw:  Make changes to the raw decoder to handle changes being made by the RAXUM team to the database.  The following database changes will be made  1) drop the quality_code column in the unkstnvalue table 2) drop columns quality_code, revision, product_id, prducttime and postingtime from table pehfsep.

       ADB Bug #:  r1-46      Programmer:  jm
        Original RFC             Entry  Date        Notes
       MBRFC                       2/23/05


Processed shefdecoder – if the shef message does not have the full SHEF code specified, the decoder gets a posting error saying that A null exists for a column that is part of the primary key.  Apparently the decoder is not fully expanding the SHEF code in the message and thus some field that is part of the primary key does not get defined.  Ex message has PPDP instead of PPDPZZZ.

        ADB Bug #:  r1-35      Programmer:  OHD/HL
        Original RFC              Entry  Date        Notes
        APRFC                        11/30/04


raw shefdecoder - the post column in the location table can have 3 values 0 = no post, 1 = post and 2 = post, see ingestfilter table.  It appears that the raw shefdecoder handles values 1 and 2 in exactly the same way.... it was intended that when the value is post =1 that the decoder insert all data type into the db ... does not need to check the ingestfilter table first.

        ADB Bug #  r1-34       Programmer: OHD/HL
        Original RFC               Entry  Date        Notes
        APRFC/MBRFC          11/30/04


Informix -- set explain sql does works, however, you get a 534 error..Reported by more than one RFC this past year.

        ADB Bug #  r1-29       Programmer: ss, jm, OHD/HL
        Original RFC               Entry  Date       Notes
        MBRFC                        11/17/03              No solution found that works at more than a couple RFCs.May be a known Informix IDS 9.3 bug.



Bugs Fixed For Release OB-7

Last Updated:  7/11/05


ivp_batch:  A memory leak is occurring in the pairing process that involves LINUX, JAVA and INFORMIX.

        ADB Bug #: r1-58    Programmer:  OHD
        Original RFC           Entry  Date        Notes
         MBRFC                    6/29/05



Bugs Fixed For Release OB-6

Last Updated:  4/28/05


shef_decode_raw – When processing archive data (5000 values  processed) the shef decoder will crash when writing to the pecrsep table.  In addition the
shef decoder has also died when running the live data feed when writing to the pecrsep table. The interim release 1 of OB5 did not have this problem.

        ADB Bug #: r1-52    Programmer:  OHD
           Original RFC         Entry  Date        Notes
         MBRFC                    3/25/05

shef_decode_pro:  Add messages to a log file(s) when ever a value to be posted is outside the time window, and thus can not be posted.  Request for OB6.

       ADB Request #:  ER-20   Programmer: ohd
        Original RFC                    Entry  Date
         MBRFC                              3/8/05

OFSSHEF: Add capability to SHEF encode OFS Snow-17 data.
 

        ADB Request #:  ER-18     Programmer: ej
        Original RFC                      Entry  Date        Build
         NCRFC/APRFC                  2/15/05

ADB Request:  Need detailed posting  information … similar to what the IHFS shefdecoder provides.  Each of the RAX decoders should have its own set of tokens for this purpose.  In the IHFS shefdecoder, the supplemental posting messages are controlled by 3 apps_defaults tokens, these are elgmess, locmess and dupmess.  In the IHFS shefdecoder these are defined as follows.

 dupmess
Specifies whether to log messages in the product log files about duplicate data, which can occur if a value is sent for a location, time, etc. for which a value already exists:
    ON   = Log messages about duplicate data.
    OFF = Don=t log messages about duplicate data.
    Default = ON

 locmess
Specifies whether to write messages in the product log file about stations and areas not defined as either a location or as an area, such as a basin, county, or zone:
    ON =   Log messages about undefined locations.
    OFF = Don=t log messages about undefined locations.
    Default = ON

 elgmess
Specifies whether to write messages in the product log files about the posting Aeligibility@ of a known location=s value.  The Aeligibility@ refers to whether the specific type of data for the given location=s data should be posted:
    ON   = Log messages about the station eligibility not being satisfied.
    OFF = Don=t log messages about the station eligibility.
    Default = ON

Messages to be considered are:
    Ignoring duplicate zzzzz  for  xxxxx  yyyyyyyy date/time
    xxxxx yyyyyyyy date/time   obs time > mm minutes in the future; data not posted
    xxxxx yyyyyyyy date/time   obs time > dd days in the past; data not posted
    xxxxx yyyyyyyy date/time   basis time > valid time; check encoding
    xxxxx not posted, station inactive
    xxxxx not defined, no data  posted
    xxxxx not defined; data posted to UnkStnValue
    xxxxx - yyyyyyyy ingest filter set to False

    xxxxx - yyyyyyyy ingest filter not defined:
         where xxxxx is location identifier(lid)
                   yyyyyyyy is shef pe code
                   zzzzz is table name
                   date/time is observation date and time of data
                   mm is the number of minutes defined by token shef_winfuture
                   dd is the numbers of days defined by token shef_winpast

        ADB Request #:  ER-9      Programmer:  OHD/HL
        Original RFC                     Entry  Date        Build
         MBRFC                               9/01/04

Provide a XMRG viewer.

        ADB Request #:  ER-2      Programmer:  al
        Original RFC                     Entry  Date        Build
         ABRFC                               1/22/04

dcextract: daily amounts from the extraction are being placed on the date that the 12Z period started, not when it ends.

       ADB Bug #:  r1-50      Programmer: ba
        Original RFC             Entry  Date        Notes
       APRFC                       3/17/05

process_stage--The application generates errors when trying to compute a flow if the location has a rating curve relationship that has negative stage values.

       ADB Bug #:  r1-40      Programmer:ba
        Original RFC               Entry  Date        Notes
      APRFC                         1/21/05

run_level1_process -- Shef files are being placed in the /rfc_arc_data/q directory instead of the /rfc_arc_data/q/processed

        ADB Bug #:  r1-36      Programmer:OHD/HL
        Original RFC               Entry  Date        Notes
         ABRFC                        12/13/04

datview - obsd time does not display correctly when listing processed data.  Also the plot of processed data appears funky.

        ADB Bug #  r1-33       Programmer: jp
        Original RFC              Entry  Date        Notes
        APRFC                         11/30/04


dcparse - its case sensitive when it reads the header card information... needs to be more flexible.

        ADB Bug #  r1-31       Programmer: ba
        Original RFC              Entry  Date        Notes
        APRFC                         11/30/04           Near Term Fix:  simply change lower case to upper case in the datacard file.

datview -- An apparent memory allocation bug in datview on the archive machine (OB1 release)  is causing the display of 15 minute to crash.

        ADB Bug #:  r1-25      Programmer: jp
        Original RFC               Entry  Date       Notes
         NCRFC                        5/24/04               RAX datview bug | Larry Ellis | 5/18/04

dcextract -- If there is an entire month of data missing from the pedpsep table (not missing data, just no entries) within the time period requested, datacard formatted output is not generated.

        ADB Bug #:  r1-22      Programmer: ba
        Original RFC               Entry  Date       Notes
         APRFC                        3/23/04

flatfile viewer: Will only show files from 2005 no matter what year is selected.  The window shows 2005 in the path even though 2003 or 2004 is selected in the setup window.

       ADB Bug #:  r1-48      Programmer:vh
        Original RFC             Entry  Date        Notes
       LMRFC/MBRFC         3/10/05

AX - ofsshef:  The resulting output files does not contain the encoded data requested in the input file.

       ADB Bug #:  r1-47      Programmer: ej
        Original RFC               Entry  Date        Notes
       NCRFC                         3/9/05


get_states: The frozen ground index, FGIX, is set to "0" in the statessacsma table, while in the input file, for the most part they are non-zero.

       ADB Bug #:  r1-49      Programmer: ba
        Original RFC             Entry  Date        Notes
       NCRFC                       3/11/05


Bugs Fixed For Release OB-5

Last Updated:  3/17/05


Provide a mechanism to facilite the writing of monthly data to a monthly table .

        ADB Request #:  ER-1      Programmer:  OHD/HL
        Original RFC                     Entry  Date        Build
         APRFC                               12/11/03

shefdecoder (raw & processed) -- Enforce the SHEF Revision code, currently both decoders ignore the "R" flag if in the SHEF message.

        ADB Bug #  r1-30       Programmer: OHD/HL
        Original RFC               Entry  Date       Notes
        MBRFC                        5/12/04
Note: Interim release for this bug is located on the anonymous ftp site (cd /oh/hsd_fixes/adb/ob5-r26.2_20050315_tar.Z

shef_decode_pro shefdecoder is not placing the max/min temperature into the correct table.  Instead of being posted to the pedpsep table it being placed in the pehpsep table.

      ADB Bug #:  r1-44      Programmer:OHD/HL
        Original RFC            Entry  Date        Notes
       APRFC                         2/8/05
Note: Interim release for this bug is located on the anonymous ftp site (cd /oh/hsd_fixes/adb/ob5-r26.2_20050315_tar.Z

shefdecoder (raw & processed) -- The decoders do NOT enforce the shef revision flag results.  This results in the latest value entered being utilized by any retrieval processes.   Need to make sure the apps handles the SHEF revision flag in the similar manner as the IHFS shefdecoder.  This will give the user the same flexibility as the IHFS shefdecoder on whether or not the SHEF revision flag is utilized.  The IHFS shefdecoder uses to apps_defaults tokens, shef_use_rev_flag and vl_always_overwrite_flag to control this.  After chatting with a HL programmer.. it should be possible to combine these tokens into one for RAX purposes. Recommend one token, with name like adb_shef_use_rev_raw where values can be as follows:

    ON = Consider the revision code when deciding whether to overwrite duplicate data.  If the revision code is not set on the new data, then the data is not replaced, unless the token vl_always_overwrite is set to ON.  The token vl_always_overwrite is still supported for this release of the SHEF Decoder, although
it is considered obsolete.

    OFF = Do not check the revision code which may be defined for the data; i.e. always overwrite duplicate data. This setting may result in unnecessary database writes, which can slow processing.  Note that despite the name of the token, a value of 1, which is typically the value for TRUE,  means do NOT use (i.e. ignore) the revision code.

Default = ON
 

        ADB Bug #:  r1-24      Programmer: OHD/HL
        Original RFC               Entry  Date        Notes
         MBRFC                        5/12/04
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z

shefdecoder_raw -- The poster for the unkstnvalue table does not always properly define the idur field of the data row that is being inserted.  This results in invalid values existing in the idur table.

        ADB Bug #:  r1-23      Programmer: OHD/HL
        Original RFC               Entry  Date        Notes
         APRFC                        5/11/04
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z

shefdecoder -- The shefdecoder(s) need to be able to write to ALL the SHEF data value tables.  The limitation in the version 1 release is not acceptable.... need to ensure these tables are written to:  commentvalue, pemrsep, pempsep, pehfseip and peqfsep.

• For pemrsep …  apps check should be:  if SHEF type code = R and SHEF dur code = M, post data to permep table.
• For pehfsep …   apps check should be:  if SHEF type code = F and SHEF dur code = H, post data to pehfsep table.
• For peqfsep …   apps check should be:  if SHEF type code = F and SHEF dur code = Q, post data to peqfsep table.
• For commentvalue .. apps check should be if data includes a SHEF encoded retained comment, post everything to the
    commentvalue but the apps should also do a check of the SHEF type and SHEF dur codes and post data to the
    appropriate SHEF data value table.

        ADB Bug #:  r1-19     Programmer: OHD/HL
        Original RFC               Entry  Date        Notes
         MBRFC                        1/13/04               This bug was reassigned to Edwin Welles 8/19/04
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z

shef_decode_process--When trying to overwrite a value with missing via SHEF messages the value(s) is not over written with missing value.

       ADB Bug #:  r1-43      Programmer:OHD/HL
        Original RFC               Entry  Date        Notes
       APRFC/NCRFC           2/7/05
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z


shef_decode_raw--When trying to overwrite a value with missing via SHEF messages the value(s) is not over written with a missing value.

       ADB Bug #:  r1-42      Programmer:OHD/HL
        Original RFC               Entry  Date        Notes
       APRFC/NCRFC            2/7/05
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z


shef_decode_raw--When the SHEF message has only a single report in it the posting summary information in the daily log file and message log file does not properly reflect that a piece of data was written to the database.

       ADB Bug #:  r1-41      Programmer:OHD/HL
        Original RFC               Entry  Date        Notes
       APRFC/MBRFC           01/25/05
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z

SHEFPARM file -- Make the SHEFPARM file the same as the SHEFPARM file used for the IHFS database.

        ADB Bug #:  r1-45      Programmer:OHD/HL
        Original RFC               Entry  Date        Notes
        NCRFC                        2/22/05
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as axrfc_ob5-r26_20050228_tar.Z


shef_decode_raw -- The raw decoder will process only 1 or 2 products and it dies.  The only thing the log file shows is that it's parsing data.

        ADB Bug #:  r1-39      Programmer:OHD/HL
        Original RFC               Entry  Date        Notes
        LMRFC                        2/2/04

Fix: Change fortran routine to stop writing to a bad unit number. Interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as shef_decode_raw.interim_20050207.Z


shef_decode_raw -- The raw decoder does not do the time check on forecast data correctly if the token adb_shef_winfuture is set to something other than 9999, forecast data does not post.  This check should work the same way as it does in the IHFS shefdecoder.

        ADB Bug #:  r1-37      Programmer: OHD/HL
        Original RFC               Entry  Date        Notes
        MBRFC                        12/10/04
Fixed with interim release placed on the anonymous ftp site in /oh/hsd_fixes/adb directory as shef_decode_raw.interim_20050207.Z.


shefdecoder_raw -- Apparently for short durations, a few of the files from the metar2shef program are being written to the raw shef decoder input directory without any permissions (i.e. as if someone had performed a chmod 000).   The shef decoder dies when trying to read these files.

        ADB Bug #:  r1-26      Programmer: OHD/HL
        Original RFC               Entry  Date       Notes
         APRFC                        5/25/04                RAX raw shef decoder hanging | Larry Ellis | 5/18/04
There were scripts added with OB5 which changed the permissions on a file.  The issue of file being written to the /rfc_arc_data directory as 000 is a DS system loading issue, and anytime an RFC fails over or loses a disk drive and go through a period of instability files may get written with 000 permissions.


dcextract -- If there is an entire month of data missing from the pedpsep table (not missing data, just no entries) within the time period requested, datacard formatted output is not generated.

       ADB Bug #:  r1-22      Programmer:ba
        Original RFC               Entry  Date        Notes
         APRFC                        3/23/04


system -- Any time files were placed in the raw shefdecoder directory with permissions of "000", the raw shefdecoder would  crash.

        ADB Bug #:  r1-27     Programmer: le
        Original RFC               Entry  Date        Notes
         NCRFC                        8/06/04

Fix:  The check script was modified to reset the permissions of any files found in the raw shefdecoder directory.


DCEXTRACT enhancements
1.  Add a verbose mode/option.  This option would display the query that is sent to the database as well as any error messages or additional information that will assist a user to determine why nothing or unexpected results are returned by the program.
2.  have apps write to the screen if no data is available
3.  use apps_defaults token for output directory.. if token value exists then it writes to the directory indicated otherwise defaults to current directory (what it does now)

        ADB Request #:  ER-8     Programmer: ba
        Original RFC                      Entry  Date        Build
         MBRFC                              5/12/04                OB5


archive database -- Tables created for API-CONT model data
       ADB Request #:  ER-11A   Programmer: jm & kh
        Original RFC                     Entry  Date        Build
         MARFC                              8/06/04                OB5

archive database -- Table created to mirror new IHFS database adjustfactor table
       ADB Request #:  ER-14A   Programmer: jm & kh
        Original RFC                      Entry  Date        Build
         APRFC                               8/06/04                OB5



Bugs Fixed For Release OB-4

Last Updated: 5/12/04


shef_decode_raw -- It has been reported that at two sites that the shef_decode_raw program is running very slowly.

        ADB Bug #:  r1-3      Programmer: jp
        Original RFC               Entry  Date        Notes
         NWRFC, ABRFC        1/8/04


shef_decode_raw -- The raw decoder stopped working when a large number of metar2shef files (>400) were placed in the /data/fxa/ispan/hydro_adbs directory.

        ADB Bug #:  r1-4      Programmer: gs
        Original RFC               Entry  Date        Notes
         ABRFC                        8/8/03


decoders -- Both the shef_decode_raw and shef_decode_pro programs unexpectedly stop.

        ADB Bug #:  r1-6      Programmer: jm
         Original RFC            Entry  Date        Notes
                                            8/20/03

Fix: Scripts have been developed for both the shef_decode_raw and shef_decode_pro programs and can be run from the cron.  A problem with the original check scripts was also reposted and the problems was resolved.  There are four (4) files available for on the anonymous ftp (ftp.nws.noaa.gov or 205.156.54.206) in the oh/hsd_fixes/adb directory:

check_processed_decoder
check_raw_decoder
adb_oper_crons
check_scripts_instuctions


datview -- When attempting to edit bad data out of the database using the datview program core dumps.  After clicking on the button to send data to the database, datview creates a zero length dvrrX product, and then proceeds to quickly memory fault.

        ADB Bug #:  r1-10      Programmer: jp
        Original RFC               Entry  Date        Notes
         ABRFC                        9/30/03


datview -- Can not adjust scale when viewing the processed data from the current 0 to 10.  It would be nice if the scale would be dynamically self adjusting as is currently available when viewing ra data.

        ADB Bug #:  r1-11      Programmer: jp
        Original RFC               Entry  Date        Notes
         APRFC                        11/19/03

dcextract -- When attempting to generate datacard format file for a location that currently has less than one year of data available, two files are produced with only header files.  dbacess shows that data does actually exist.

        ADB Bug #:  r1-12      Programmer: ba
        Original RFC               Entry  Date        Notes
         APRFC                        11/19/03

Fix:  The needed file is on the anonymous ftp server (205.156.54.206).  You can retrieve the needed files on the anonymous ftp server, after logging on, by typing in the following commands:

    cd hsd_fixes/adb
    binary
    get dcextract.gz
    quit


level one processors -- Can not load data older than 1990 into the processed database using the level one processors.

        ADB Bug #:  r1-13     Programmer: ba/jm
        Original RFC               Entry  Date        Notes
         APRFC                        9/30/03

Fix:  The needed file is on the anonymous ftp server (205.156.54.206) and are the same files that are needed to fix bug r1-14.  You can retrieve the needed files on the anonymous ftp server, after logging on, by typing in the following commands:

    cd hsd_fixes/adb
    binary
    get process.tar.gz
    get transfer.tar.gz
    get cbrfc.a.gz
    get
    quit


level one processors -- Can not push more than 30 days of data into the processed data tables using the level one processors.

        ADB Bug #:  r1-14      Programmer: ba/jm
        Original RFC               Entry  Date        Notes
         APRFC                        11/19/03

Fix:  The needed files are on the anonymous ftp server (205.156.54.206) and are the same files that are needed to fix bug r1-13.  You can retrieve the needed files on the anonymous ftp server, after logging on, by typing in the following commands:

    cd hsd_fixes/adb
    binary
    get process.tar.gz
    get transfer.tar.gz
    get cbrfc.a.gz
    get
    quit


datview -- When requesting text and plotted data, the plotted data appears but the text data does not.

        ADB Bug #:  r1-16     Programmer: mt
        Original RFC               Entry  Date        Notes
         NWRFC                       7/30/03


process_precip -- The level 1 processor, process_precip does not seem to be transforming the SHEF PE "PC" properly.  It appears to be ignoring values and calculating missing too often for the various totals it computes.

        ADB Bug #:  r1-17      Programmer: ba
        Original RFC               Entry  Date        Notes
         NCRFC, APRFC         1/13/04


adbinit -- when following the instructions for the adbinit suite discovered the run_table scripts that was supplosed to be in the /rfc_arc/bin/adbinit/hdb directory was missing.

        ADB Bug #:  r1-21      Programmer: jm
        Original RFC               Entry  Date        Notes
         CBRFC                        1/26/04

Fix:  The needed file is on the anonymous ftp server (205.156.54.206).  You can retrieve the needed files on the anonymous ftp server, after logging on, by typing in the following commands:

    cd hsd_fixes/adb
    binary
    get rtfix.tar.gz
    quit

The documentation for this fix will replace the existing online documentation located on the RFC archive database web page.  If you have any problems or questions, contact the RFC Support Team.



Bugs Fixed For Release OB-3

Last Updated: 5/12/04


ISQL -- The ISQL  program that came with the version of RedHat Linux delivered on the archive machines.  It is in the /usr/bin directory and will come up if you type isql at the command prompt if /usr/bin is in the oper PATH before the /opt/informix/bin directory.

        ADB Bug #:  r1-1      Programmer: jm
        Original RFC             Entry  Date        Notes
                                            8/1/03

Fix: Section E (Configure the Archive Server) Step 3 of the Mod Note has some code for updating the /home/oper/.profile file.  This file may not look the same in all offices so Julie Meyer (MBRFC) has contributed her additions to the oper .profile file.  They work for Julie using a dtterm window.  The may be found on the anonymous ftp  server in the hsd_fixes/adb directory.  The file is called profile_additions.txt.


Post Install Instructions -- Incorrect load sequence in the post install instructions.

        ADB Bug #:  r1-2      Programmer: jm
         Original RFC            Entry Date        Notes
                                            8/1/03

Fix: Updated files are available for on the anonymous ftp (ftp.nws.noaa.gov or 205.156.54.206) in the oh/hsd_fixes/adb directory.  Three (3) files needed to down load:

cmd_ld-xref
shefpetrans-all.tar.Z
shefpetrans-patch-instructions.wpd


metar2shef -- The metar2shef script neither copies nor removes files if more than a certain number, a few hundred, METARS are processed during a single run of "run_metar2shef".

        ADB Bug #:  r1-5      Programmer: ji
        Original RFC             Entry  Date        Notes
         ABRFC                        8/8/03

Fix:  Changes were made to the metar2shef script and will be included in the OB3 release.


Data Transfer (PART 1) -- Data is not being transferred into the processed tables.  The inputparm file used by the SHEF decoders is missing the SHEF TS codes used for the processed data.

        ADB Bug #:  r1-7      Programmer: jm
         Original RFC            Entry Date        Notes
                                            8/20/03

Fix: A new inputparm file is available on the anonymous ftp (ftp.nws.noaa.gov or 205.156.54.206) in the oh/hsd_fixes/adb directory.  It is called inputparm-new.  It should be placed in the /rfc_arc/cfg/decoders directory and renamed inputparm (after backing up the original).  You will need to restart the decoders.


Data Transfer (PART 2) -- Data is not being transferred into the processed tables.

        ADB Bug #:  r1-8      Programmer: jm
         Original RFC            Entry  Date        Notes
                                            8/20/03

Fix: If the user has not defined all SHEF PE codes with corresponding level 1 SHEF type/souce codes, then the shef_decode_pro decoder needs to be started with a "-m" option on the command line.  This will bypass the checking of the location, ingestfilter and aliasid tables.  Also, since they are run from the cron, the scripts to start and run the
processed decoder need some environment variables defined.  The new version of these scripts is available on the anonymous ftp (ftp.nws.noaa.gov or 205.156.54.206) in the oh/hsd_fixes/adb directory.  There are two (2) files:

run_processed_decoder
start_processed_decoder

The files should be placed in the /rfc_arc/scripts/decoders directory.  Please make sure they're executable.


Post Install Instructions -- Removing unnecessary tables.

        ADB Bug #:  r1-9      Programmer: jm
         Original RFC              Entry  Date        Notes
                                              8/20/03

Fix: There were 2 unneeded tables delivered - you can drop the following tables by using dbaccess.  The two (2) tables are:

counties_dia
counties_vio

Please use the following instructions to drop the two tables mentioned above:
1.  at prompt: dbaccess [Enter]
2.  press [T] for table
3.  database name should be highlighted, adb_ob1xxx@adbs, press [Enter]
4.  ring menu for Table option should now be displayed.
5.  press [D] for Drop
6.  at list of table names will appear, using the up/down arrow keys highlight the table name, the user wishes to drop,

   counties_dia and press [Enter]

7.  a question asking you to confirm, highlight Yes if you are sure you selected the correct table name
8.  screen will indicated table was dropped and return user to the Table ring menu, repeat steps 5-7 except this time the table name is counies_vio
9.  press [E] key twice to return to the command line prompt


Enhancement Requests-Unassigned

Last Updated:  4/28/05
New Additions to Latest Update:  ER-30


shef_decode_pro:  add option to check the aliased table if the lid is not found in the location table.  Then use the alternative lid if one is found in the aliasid table and recheck the location table to determine if data should be posted to archive db.  This feature was supposedly available in the ob1 version of the software.

       ADB Request #: ER-30    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05

shef_decode_raw:  add option to check the aliased table if the lid is not found in the location table.  Then use the alternative lid if one is found in the aliasid table and recheck the location table to determine if data should be posted to archive db.  This feature was supposedly available in the ob1 version of the software.

       ADB Request #: ER-29    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05


shef_decode_pro: Misc code clean-up suggested by HL programmers.

       ADB Request #: ER-28    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05


shef_decode_raw: Misc code clean-up suggested by HL programmers.

       ADB Request #: ER-27    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05


shef_decode_pro: Add counters to count numbers for data not posted.  This information would be part of the posting information that is written to both the daily log file and message output file.  The counters would be
1) IgnorLoc – counts data not posted because post column in location table is set to zero (no post for that location identifier (lid)
2) IgnorIng – counts data not posted because ingest column in ingestfilter table is set to zero (no post) for that lid and SHEF code
3) IgnorDup – counts data already in db and SHEF revision flag is not on .A, .B, or .E line
4) IgnorUnk – token adb_shef_raw_post_unk is set to NONE, counts data not posted because lid is not in location table
There counters are similar to the counters that exists in the operational IHFS shefdecoder and should function in a similar manner.
Suggested format for line to be written to logfiles:
Not Posted:  IgnorLoc -    0   IgnorIng -      0   IgnorUnk -     0   InnorDup –

       ADB Request #: ER-26    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05

shef_decode_raw:  Add counters to count numbers for data not posted.  This information would be part of the posting information that is written to both the daily log file and message output file.  The counters would be
1) IgnorLoc – counts data not posted because post column in location table is set to zero (no post for that location identifier (lid)
2) IgnorIng – counts data not posted because ingest column in ingestfilter table is set to zero (no post) for that lid and SHEF code
3) IgnorDup – counts data already in db and SHEF revision flag is not on .A, .B, or .E line
4) IgnorUnk – token adb_shef_raw_post_unk is set to NONE, counts data not posted because lid is not in location table
There counters are similar to the counters that exists in the operational IHFS shefdecoder and should function in a similar manner.
Suggested format for line to be written to logfiles:
Not Posted:  IgnorLoc -    0   IgnorIng -      0   IgnorUnk -     0   InnorDup –

       ADB Request #: ER-25    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05

shef_decode_pro:  Implement locdatalimits/datalimits checks.

       ADB Request #: ER-24    Programmer:
        Original RFC                    Entry  Date        Notes
       MBRFC                              4/25/05

transfer_precip--Add ability to handle precip data coded as PPP that have an observation time plus/minus 2hrs around 7am local (12Z).  The apps should switch the duration code from "P" to "D" when it creates the SHEF message for the processed shefdecoder.

        ADB Request #: ER-23    Programmer:
        Original RFC                     Entry  Date        Build
         MBRFC                              3/30/05

process_stage--Add an option to have a choice on if it SHEF encodes both stage and flow, stage only, or flow only.

        ADB Request #: ER-17    Programmer:
        Original RFC                     Entry  Date        Build
         MBRFC                              1/21/05

Datview--Need a new option when editing stage data in one of the processed value data tables.  This option would allow the user to have a flow value calculated and encoded in SHEF along with the edited stage value.

        ADB Request #: ER-16    Programmer:
        Original RFC                     Entry  Date        Build
         APRFC                               1/21/05


raw shefdecoder -  need feature similar to operational shefdecoderwhere if post flag says to post in location table and SHEF codeis not in ingestfilter that the shefdecoder inserts the row into ingestfilter automatically.   Since all RFCs may not want this option it needs to be controlled the same way it is in the operational shefdecoder; with an apps defaults token.

        ADB Request #: ER-15    Programmer:
        Original RFC                     Entry  Date        Build
         APRFC                               12/06/04


Make necessary application changes to raw shefdecoder to utilize the new adjustfactor table.
Note:table was added in build OB5 under enhancement ER-14A

        ADB Request #: ER-14    Programmer:
        Original RFC                     Entry  Date        Build
         APRFC                              8/06/04


Browser based viewer for the flatfiles archive

        ADB Request #: ER-13    Programmer:
        Original RFC                     Entry  Date        Build
         NWRFC                             5/12/04


GUI application to access the meta-data and quality control tables..  Something like a stripped down version of the WHFS HYDROBASE application.  Application needs to some how see selected tables in the IHFS database (ie., location, ingestfilter, riverstat, etc)

        ADB Request #: ER-12     Programmer:
        Original RFC                      Entry  Date        Build
         MBRFC                               5/12/04


Request new db NWSRFS tables for API-CONT model data and application(s) that will read/write OFS information to these tables
Note:table was added in build OB5 under enhancement ER-11A
        ADB Request #: ER-11   Programmer:
        Original RFC                     Entry  Date        Build
         MARFC                              5/12/04

DATVIEW enhancement -We need to be able to quickly look at historical data, see how it compares to climate, be able to easily adjust the window that we are looking at, be able to edit any bad data that slipped through the level 1 processors.We would like to be able to look through a year’s worth of data for a particular station and type source, and look at average daily, max daily, min daily for z day or for local day.

        ADB Request #: ER-10    Programmer:
        Original RFC                      Entry  Date        Build
        APRFC                                 5/12/04


Specific hydrograph plotting capability:
1.  User-selectable location, parameter (stage or streamflow), time frame, and minimum value for peak flow.
2.  All hydrographs meeting the criteria are displayed on the same plot.
3.  Time scale of each hydrograph is adjusted so that peaks coincide.
4.  User has capability to remove any hydrograph from the group.
5.  Capable of deriving a representative unit hydrograph from the accepted storm (streamflow) hydrographs.

        ADB Request #: ER-5      Programmer:
        Original RFC                     Entry  Date        Build
         MBRFC                               5/12/04


A graphical data display shall be developed to display spatial and temporal data from an archived database.  Forecasters shall be able to select stage, precipitation and temperature data for multiple sites for any time period archived and quickly display plots of the parameter with time.  Plots of stage shall have the capability of plotting multiple gaging stations on the same graph.  In addition, the display program shall be capable of spatially displaying all values for a particular parameter at any given time.

        ADB Request #: ER-3      Programmer:
        Original RFC                     Entry  Date        Build
         MBRFC                               5/12/04


Enhancement Requests-Assigned

Last Updated:  4/28/05



datview - there appears to be a limitation on the earliest date that can be selected... i.e.1970.   Need to have the date field be more flexible ... go back to at least 1952 if not earlier.

        ADB Bug #  r1-32       Programmer: jp
        Original RFC              Entry  Date        Notes
        APRFC                         11/30/04

arcmenu:  Add new submenu for start/stop scripts of the shef decoders

       ADB Request #: ER-22    Programmer: jm
        Original RFC                    Entry  Date        Notes
       MBRFC                             3/28/05


loadmods/fetchmods:  Allow the programs to be able to handle the following mods: APICQN, BFRATE, CBASEF, CBFRATE, UCBASEF, UCBFRATE, UHGCDATE, and WEUPDATE. In addition to the software changes, this will require eight new tables to be added to the archive db.

       ADB Request #: ER-21    Programmer:ba
        Original RFC                   Entry  Date        Notes
       NCRFC/MBRFC                3/10/05


loadmods/fetchmods:  Allow the programs to be able to handle mods which have comments in them and also accept the use local time in the mods. Loadmods/fetchmods needs to be able to handle the mods BASEF and BFRCHNG.  ADB will need to have tables for the apps to write to and read from.

       ADB Request #: ER-19   Programmer:ba
        Original RFC                    Entry  Date        Notes
       NCRFC                              3/3/05


Areal Plotting Capabilities:
1.  Areal display of data allowing point or areal averaged values displayed for same time period or window of time.  Provide color coding based on thresholds similar to XNAV.
2.  The whole display capability should be similar to XNAV but have capability to plot parameters other than height in a time series and have more capability in the plots to adjust scales and display more than one plot at a time either on the same plot or in separate plots.

        ADB Request #: ER-6      Programmer:  al
        Original RFC                     Entry  Date        Build
         MBRFC                               5/12/04


DATVIEW enhancements:
1.  Allow both “observed” and forecast time series  to be plotted on the same plot for the selected station.
2.  Allow up to 3 stations with compatible PE data to be plotted
3.  Allow multiple compatible PE’s from a single station to be plotted.  Example: a river reach of locks and dams with hinge-point control: user wishes  to plot the downstream control (HP), the control point (HG), and the upstream inflow (HT).  Another example is plotting QR from the control point and QT from the upstream and downstream projects.  A third     example is plotting both HP and HT from the lock and dam site.
4.  Implement option to interactively apply the zero    datum(s) to all or some of the specified lid/PE pairs,  to convert displayed or output data to a common datum (MSL-1912, NGVD 1929, NAVD-88, etc.)   Note that at some projects, the HP and HT have different     datums, such as HP  in MSL while HT is referenced to an arbitrary datum.  Thus a single “apply zero datum” toggle to apply zero datums to all lid/PE pairs displayed  would be undesirable.

        ADB Request #: ER-7      Programmer: jp
        Original RFC                     Entry  Date        Build
         MBRFC                               5/12/04

Time series plotting capabilities with the following general features:
1.  User-selectable locations/areas/parameters.
2.  Easy capability to adjust time scale to shorten or expand time frame in window.
3.  Capable of adjusting y axis scale so as to provide opportunity to ignore extreme data values.
4.  Capable of plotting multiple time series on same plot as long as they are same parameter or same type of units.
5.  Capable of displaying multiple plot windows at a time so that comparisons can be made between plots of different IDs, or plots of the same parameter and ID but of different time frame.
6.  Capable of plotting accumulation of values for a window of time such as precip or snow.

        ADB Request #: ER-4      Programmer:  jp
        Original RFC                     Entry  Date        Build
         MBRFC                               5/12/04



Initial System Fixes

Additional maintenance system fixes will be documented in the appropriate OB Release section
Last Updated:  08/23/04


system .Apps_defaults tokens -- There are the 2 tokens that are incorrectly defined in the .Apps_defaults file in the rax's /awips/hydroapps directory.
Correct deifinitions are shown below:

            adb_shef_pro_logs_dir  :  $(adb_logs_dir)/decoder/processed/logs
            adb_shef_pro_err_dir  :  $(adb_logs_dir)/decoder/processed/err

        ADB Bug #:  r1-28      Programmer: jm
        Original RFC               Entry  Date        Notes
         MBRFC                        8/23/04

Fix:  Check the /awips/hydroapps/.Apps_defaults file on your rax system and correct the token(s) if necessary.


system/password -- Since the initial install the passwords for root and informix have been changed more than once.   Changing the passwords for these users has somehow created a problem connecting to the Informix IDS engine when trying to use the Informix tools dbaccess and isql.  A -951 error occurs when trying to select a database.  Testing run at APRFC and NCRFC, where the passwords for root and informix have not been changed from the initial install, did not duplicate this problem.  The only item of interest noted by both APRFC and NCRFC was they had to login in as root on the ds first then go to the rax to run dbaccess and isql successfully.  Others in the field with Informix expertise have been consulted since this problem was originally discovered back in mid November with no success at a solution.  This problem needs to be elevated to above the RAXUM team.

        ADB Bug #:  r1-18      Programmer:  rr
        Original RFC               Entry  Date        Notes
         MBRFC                       1/13/04

Fix:  When the password was changed, the processed created a /etc/shadow file.  Since the password shadowing is turned OFF  this file should not be present.  To fix this problem, run /usr/sbin/pwunconv.  This creates a new passwd file from the passwd file and the shadow file and then removes the shadow file.


system -- The linux operating system does not have the updated versions of the dump and restore commands as stated in the user documentation:  "The following scripts assume that the version of the dump and restore commands on the Linux system are dump-0.4b27-3 or later".
The installing of these updated commands should have been part of the install process.

        ADB Bug #:  r1-20      Programmer:  rr
        Original RFC               Entry  Date        Notes
         MBRFC                        1/13/04

Fix:  An RPM file has been obtained and instructions for installing the new RPM have been written.  When you are ready to install the RPM contact Randy Rieman at the RFC Support Group.