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

[IDD #BKI-822215]: LDM feed sites



Hi Adam,

re:
> Ok, my ldmd.conf request lines are as follows:
> 
> #
> request UNIDATA|NIMAGE|FSL2|FNEXRAD     ".*"    pavan.srcc.lsu.edu PRIMARY
> request UNIDATA|NIMAGE|FSL2|FNEXRAD     ".*"    idd.cise-nsf.gov SECONDARY
> #
> request NGRID   ".*"    pavan.srcc.lsu.edu      PRIMARY
> request NGRID   ".*"    idd.cise-nsf.gov        SECONDARY
> #
> request NEXRAD3 ".*"    pavan.srcc.lsu.edu      PRIMARY
> request NEXRAD3 ".*"    idd.cise-nsf.gov        SECONDARY
> #
> request NEXRAD2 ".*"    idd.cise-nsf.gov        PRIMARY
> request NEXRAD2 ".*"    idd.unidata.ucar.edu    SECONDARY
> #
> request FNMOC|GEM       ".*"    idd.cise-nsf.gov        PRIMARY
> request FNMOC|GEM       ".*"    idd.unidata.ucar.edu    SECONDARY
> #
> request CONDUIT "([09]$)" idd.cise-nsf.gov      PRIMARY
> request CONDUIT "([18]$)" idd.cise-nsf.gov      PRIMARY
> request CONDUIT "([27]$)" idd.cise-nsf.gov      PRIMARY
> request CONDUIT "([36]$)" idd.cise-nsf.gov      PRIMARY
> request CONDUIT "([45]$)" idd.cise-nsf.gov      PRIMARY
> #
> request CONDUIT "([09]$)" idd.unidata.ucar.edu  SECONDARY
> request CONDUIT "([18]$)" idd.unidata.ucar.edu  SECONDARY
> request CONDUIT "([27]$)" idd.unidata.ucar.edu  SECONDARY
> request CONDUIT "([36]$)" idd.unidata.ucar.edu  SECONDARY
> request CONDUIT "([45]$)" idd.unidata.ucar.edu  SECONDARY
> #
> # NLDN Feeds
> #
> 
> Does this look correct?

Yes.  FYI: the PRIMARY and SECONDARY designations no longer
mean anything; they can be removed.  I personally think that
their meaning should be reinstated, but this is an argument
that I have not won.  If they do get reinstated at some point
down the road, there would be three modes:

PRIMARY
SECONDARY
AUTO

The default action now is AUTO meaning that the feed can
automatically be promoted from SECONDARY to PRIMARY or
demoted from PRIMARY to SECONDARY based on how many products
is received from each upstream in a fixed amount of time.

> stokes.metr.ou.edu is listed as my secondary but
> notifyme -h stokes.metr.ou.edu shows me nothing and give me this in the
> logs:
> 
> request NLDN     ".*"    striker2.atmos.albany.edu Jul 01 19:55:58 tornado
> notifyme[12032] NOTE: Starting Up: stokes.metr.ou.edu: 20090701195558.954
> TS_ENDT {{ANY,  ".*"}}
> Jul 01 19:55:58 tornado notifyme[12032] NOTE: LDM-5 desired product-class:
> 20090701195558.954 TS_ENDT {{ANY,  ".*"}}
> Jul 01 19:55:59 tornado notifyme[12032] NOTE: NOTIFYME(stokes.metr.ou.edu):
> reclass: 20090701195558.954 TS_ENDT {{UNIDATA,  ".*"}}
> Jul 01 19:55:59 tornado notifyme[12032] NOTE: NOTIFYME(stokes.metr.ou.edu):
> OK
> #

OU has been struggling for quite a long time on their ability to relay
IDD data.

> Also...is it normal when a PRIMARY/SECONDARY entry is being used for it to
> bounce around so much:
> 
> Jul 01 19:57:46 tornado pavan.srcc.lsu.edu[11594] NOTE: Switching
> data-product transfer-mode to alternate
> Jul 01 19:57:46 tornado pavan.srcc.lsu.edu[11594] NOTE: LDM-6 desired
> product-class: 20090701185746.365 TS_ENDT {{NGRID,  ".*"},{NONE,
> "SIG=5f54fe5e37ac3347865ddf59c9a5c0e0"}}
> Jul 01 19:57:46 tornado pavan.srcc.lsu.edu[11594] NOTE: Upstream LDM-6 on
> pavan.srcc.lsu.edu is willing to be an alternate feeder
> Jul 01 19:57:46 tornado idd.cise-nsf.gov[11595] NOTE: Switching data-product
> transfer-mode to primary
> Jul 01 19:57:46 tornado idd.cise-nsf.gov[11595] NOTE: LDM-6 desired
> product-class: 20090701185746.983 TS_ENDT {{NGRID,  ".*"},{NONE,
> "SIG=2f6deaf1b817f2e3a67f9211bdbecf5e"}}
> Jul 01 19:57:47 tornado idd.cise-nsf.gov[11595] NOTE: Upstream LDM-6 on
> idd.cise-nsf.gov is willing to be a primary feeder
> Jul 01 19:59:05 tornado idd.cise-nsf.gov[11595] NOTE: Switching data-product
> transfer-mode to alternate
> Jul 01 19:59:05 tornado idd.cise-nsf.gov[11595] NOTE: LDM-6 desired
> product-class: 20090701185905.514 TS_ENDT {{NGRID,  ".*"},{NONE,
> "SIG=c6233bffa9404bd6b9c70268f7b1c6d4"}}
> Jul 01 19:59:05 tornado idd.cise-nsf.gov[11595] NOTE: Upstream LDM-6 on
> idd.cise-nsf.gov is willing to be an alternate feeder
> Jul 01 19:59:06 tornado pavan.srcc.lsu.edu[11594] NOTE: Switching
> data-product transfer-mode to primary
> Jul 01 19:59:06 tornado pavan.srcc.lsu.edu[11594] NOTE: LDM-6 desired
> product-class: 20090701185906.019 TS_ENDT {{NGRID,  ".*"},{NONE,
> "SIG=c6233bffa9404bd6b9c70268f7b1c6d4"}}
> Jul 01 19:59:06 tornado pavan.srcc.lsu.edu[11594] NOTE: Upstream LDM-6 on
> pavan.srcc.lsu.edu is willing to be a primary feeder
> 
> This is happening for every feed in which there is an PRIMARY/SECONDARY
> option...

This is normal when the latency for the two upstream sites is very
close to being the same.  FYI: you can force the feed requests to
always be primary by making the request patterns distinct.  Here is
the example for the feeds from pavan and idd.cise-nsf.gov:

request UNIDATA|NIMAGE|FSL2|FNEXRAD ".*" pavan.srcc.lsu.edu
request UNIDATA|NIMAGE|FSL2|FNEXRAD "(.*)" idd.cise-nsf.gov

The patterns are effectively the same, but they are seen as being
different in the receiving processes.  Since they are different,
the data will be sent in PRIMARY mode on both connections.  This
will, of course, use twice the incoming bandwidth.  The job of
elimination of duplicate products then falls to the queue insertion
logic: if a product with the same MD5 signature is already in
the queue, the one just received will be discarded.

> Thanks,

No worries.

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: BKI-822215
Department: Support IDD
Priority: Normal
Status: Closed