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

20030805: IDD Changes at LSU



>From: Robert Leche <address@hidden>
>Organization: LSU/SRCC
>Keywords: 200308052029.h75KTKLd000765 LDM IDD

Hi Bob,

>LSU's IDD ingestor Sirocco has shown signs of instability as of late. 
>The LSU downstream feeders have already moved to Seistan, I believe now 
>is the time to make Seistan LSU's primary IDD feeder as well.

Sounds reasonable.

>Currently 
>Sirocco is feeding from Datoo.srcc.lsu.edu, Thelma.ucar.edu, and 
>Unidata2.ssec.wisc.edu with the following request configuration:
>#
>request         NEXRAD|IDS|UNIDATA|HRS|WSI|WMO|DDS|HDS|FNEXRAD|NPORT ".*"    
>datoo.srcc.lsu.edu
>request         DIFAX|FSL2|MCIDAS  ".*"    thelma.ucar.edu 
>request         IDS|DDPLUS  ".*" unidata2.ssec.wisc.edu 
>request         HRS         ".*" unidata2.ssec.wisc.edu 
>request         UNIWISC     ".*" unidata2.ssec.wisc.edu 
>request         NNEXRAD     ".*" unidata2.ssec.wisc.edu 
>request         NMC3        ".*" datoo.srcc.lsu.edu


>Seistan's request configuration would change to:


># IDS|DDPLUS - world observational data from NOAAPORT
>#
>request IDS|DDPLUS      ".*"    datoo.srcc.lsu.edu      PRIMARY
>request IDS|DDPLUS      ".*"    unidata2.srcc.lsu.edu   ALTERNATE
>#request IDS|DDPLUS      ".*"    sirocco.srcc.lsu.edu    ALTERNATE

This looks good.  If sirocco is going away for good, you should
remove the request line to it in order to keep seistan's ldmd.conf
file readable.

>#
># HDS - model output from NOAAPORT
>#
>request HDS     ".*"    datoo.srcc.lsu.edu      PRIMARY
>request HDS     ".*"    thelma.ucar.edu         ALTERNATE
>#request        HDS     ".*"    sirocco.srcc.lsu.edu    ALTERNATE

Please change thelma.ucar.edu to atm.geo.nsf.gov.  This is part of
the reorganization of the IDD.  Ditto the comment about removing
the request to sirocco.

>#
># NNEXRAD    - NEXRAD Level III products from NOAAPORT
>#
>request NNEXRAD ".*"    datoo.srcc.lsu.edu      PRIMARY
>request NNEXRAD ".*"    unidata2.srcc.lsu.edu   ALTERNATE
>#request        NNEXRAD ".*"    sirocco.srcc.lsu.edu    ALTERNATE

This looks good.  Ditto the comment about removing the request to
sirocco.

>#
># FNEXRAD - NEXRAD Level III composite produts
># UNIWISC - Unidata-Wisconsin GOES satellite imagery and products
>#
>request FNEXRAD|UNIWISC ".*"    unidata2.srcc.lsu.edu           PRIMARY
>#request        FNEXRAD|UNIWISC ".*"    sirocco.srcc.lsu.edu    PRIMARY

Ditto the comment about removing the request to sirocco.

Please add:

request FNEXRAD|UNIWISC ".*"    atm.geo.nsf.gov    ALTERNATE

># DIFAX - DIFAX-like products created at University of Wisconsin/AOS
># FSL2  - NOAA/FSL wind profiler data
>#
>request DIFAX|FSL2|MCIDAS       ".*"    thelma.ucar.edu         PRIMARY
>#request        DIFAX|FSL2      ".*"    sirocco.srcc.lsu.edu    PRIMARY

Please change thelma.ucar.edu to atm.geo.nsf.gov.  This is part of the
reorganization of the IDD.  Ditto the comment about removing the
request to sirocco.

>request         NMC3            ".*"    datoo.srcc.lsu.edu      PRIMARY
>request         HRS             ".*"    unidata2.srcc.lsu.edu   PRIMARY
>request         WSI             ".*"    datoo.srcc.lsu.edu      PRIMARY
>request         WMO             ".*"    datoo.srcc.lsu.edu      PRIMARY
>request         DDS             ".*"    datoo.srcc.lsu.edu      PRIMARY
>request         NPORT           ".*"    datoo.srcc.lsu.edu      PRIMARY

This list should be changed.  Here are some comments:

- NMC3 is FNEXRAD.

  Unless datoo is requesting FNEXRAD or WSI from an upstream
  machine -- and it should not be --, having these requests will simply
  result in more rpc.ldmds running, but they won't be doing anything
  but using process IDs and CPU time.

- WMO is the combination of IDS|DDPLUS|HDS; DDPLUS is the combination
  of DDS|PPS.  Given this, the requests for HRS, DDS and WMO overlap.
  Not only that, you already have all feeds in WMO covered by the
  two requests in the IDS|DDPLUS and HDS sections above.

- NPORT is not used

Given these last items, _all_ of the above requests should be dumped.

>I believe the current settings on Thelma and Unidata2.ssec.wisc.edu 
>will allow Seistan to feed.

Thelma is not setup to feed LSU, but atm is.

>This being the case, the plan is to direct 
>Seistan to feed directly off of Thelma and  unidata2.ssec.wisc.edu 
>taking Sirocco out of the loop.

It should feed off of datoo, unidata2, and atm.

>After a short proving period I will then 
>direct Sirocco to feed from Seistan locally.

OK.

Tom