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

[Datastream #AYJ-610338]: Level 2 radar delay intentional?



Hi Gilbert,

re: the latencies indicated at your site are on the order of a second
> Right, but...

re: what the time in the Level II header really is
> Ahhh. Can you verify this for me? This is the 7 minute delay I am talking
> about. 

The time in the product and product header should be the nominal time for
the scan.

> Still, when I want to make an image, I get the full level 3 image
> before the level 2 image by about 7 minutes.

Evidently the Level III products are produced at the PUP while the
volume scan is still in progress.

re: a full volume scan takes on the order of 5 minutes to complete
> OK.

re: use  notifyme -vxl- -f NEXRAD2 -h toplevel_LevelII_relay
> Yep...

re: use the realtime stats pages to work back through the relay sequence
> OK, but I do not know the names of the machines feeding me data.
> Obviously, it's your machine, but from there...?

The missing link here is the names of the data server nodes of the
idd.unidata.ucar.edu cluster.  The machines are uni1, uni2, and uni3.
Exactly which one is serving you data can be semi-dynamic (persistent
after a connection is established; dynamic when connecting a new request).
All three cluster data servers get their data from the cluster collector
machines, oliver and emo.  So, you can look at the realtime stats for
either/both oliver and emo and see who they are feeding from.  The
latency seen will be the _minimum_ from either of these two machines.
You can trace who they are feeding from using the real time stats
'topology' links for NEXRAD2 for them.

re: there are still some NEXRADs that have improperly maintained clocks
> Actually, I was wondering if that was the problem as I read through your
> email! If this is the problem, then if these are Linux boxes, activating
> the NTP service is a snap and will fix the problem immediately.
> 
> Thoughts?

The time _in_ the product will be correct since it is set with a GPS
timeserver.  The indicated time that the product was created will be incorrect
if the clock on the LDM machine is incorrect.

Posted on 20060309.2226 MST     

> BTW, I'd be happy to let them use my NTP server...

Access to a timeserver is not the problem at the NEXRADs with the incorrect
clocks.  The procedures for getting the LDM injection machine upgraded
is the limiting factor (i.e., upgrades are done by the NWS in "builds").
There is _nothing_ that we can do about this other than to continue to
recommend that they get all clocks synchronized.  Believe me when I tell
you that we have been making this same recommendation since day 1.

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: AYJ-610338
Department: Support Datastream
Priority: Normal
Status: Closed