[Ncep.nhc.nco_contacts] TSB Morning Rounds - Thu Sep 22, 2016
nhc.tsbadmin at noaa.gov
Thu Sep 22 15:32:32 UTC 2016
Here is today's summary of NHC computer operations:
--- Craig Mattocks
1. GFE/NCP issues: NCF (AWIPS Network Control Facility - user support)
restarted the AWIPS Linux Data Server DX6 for TAFB forcasters.
2. The Pacific/NH1 NWPS wave model data was not coming in, very fragmented,
since yesterday afternoon. This impacts the Wave Height, Swell, Period,
etc. TSB will investigate these problems, which were probably caused by the
failure of the NWS CP socket server (see Other Items below).
3. N-AWIPS issues: The video card on the Starfish workstation failed upon
rebooting. The reboot messages scrolled down like usual but the screen went
blank. At 1942 UTC, TSB Developer Chris Mello reported this issue to the
SDM, who submitted a NCO Help Desk ticket and followed up with a call. It
seems that recent updates to the Linux kernel and other software packages,
especially to the Xorg/X11 Window packages, required a different driver to
be installed. At 2010 UTC, NCO reported that the system was functioning
properly again. However, many other operational Linux systems now require
the installation of different NVIDIA video drivers. NCO Linux SysAdmin Curt
Steinmetz ran an inventory and is developing a plan to modify the drivers
on these machines:
Hostname Video Card Current Driver Good Driver
crappie K600 346.72 304.131
sheephead K600 346.72 304.131
hurrback NVS310 346.72 304.131
ridges Quadro 2000 346.72 367.35
tilefish K620 346.72 367.45.11
sawfish K600 346.72 304.131
tuna K620 346.72 367.45.11
bonefish NVS310 346.72 304.131
cumulus GTX570 346.72 ???
starfish K600 375.35 304.131
ballyhoo K620 346.72 367.45.11
grunt K620 346.72 367.45.11
shiner K600 361.42 304.131
4. Two light fixtures over the TAFB backup desk have been inoperable for
months. After the problem was first reported all of the light bulbs were
swapped out for the new kind which have since been swapped back, however,
the two broken fixtures never were actually repaired. This issue was
reported to Ms. Jannette Perez on Wednesday, September 21, who informed
maintenance about the situation.
5. New ocean wave prediction capability. TSB AWIPS SysAdmin Chris Mello has
implemented the ingestion and display of output from the new experimental
RWPS (Regional Wave Prediction System) in GFE. The RWPS uses the Wavewatch
III (ww3) model instead of the SWAN nearshore wave model. This system runs
on the WCOSS supercomputing clusters and includes a parameterization for
swell dissipation, which should improve wave simulations over vast oceanic
areas like the TAFB Pacific domain. The simulations have not had time to
develop "hot start" initial conditions and developers in EMC's Wave
Modeling Team are still working on including surface current forcing, so
forecasts of wave amplitude are currently too low, but forecast quality
should improve in the near future.
1. N-AWIPS/AWIPS2 issues: TSB requested that HSU not reboot any NAWIPS
boxes since there may be an issue with the video driver due to the latest
patch that was installed. If you have NAWIPS workstation issues, please
call the SDM.
2. Other Techincal issues: The text TWOs (Tropical Weather Outlooks) did
not update on the webpage at 18Z. The GTWO (Graphical Tropical Weather
Outlook) ran fine. TSB Web Developer Dave Zelinsky informed HSU that there
were issues with dataflow caused by issues outside of NHC (socket server
failures - see Other items below).
3. N-AWIPS/AWIPS2 issues: High-resolution ASCAT-A and ambiguities have not
arrived since about 21Z yesterday. Low resolution data available and ASCAT
B data coming in. This is similar to what occurred on Monday. TSB will
4. NCO began a new 24-hour parallel production test on the Gyre and Surge
WCOSS supercomputing clusters starting with the 12Z cycle today. TSB
Developers Monica Bozeman and Matt Sardi modified NHC's 'prod2dev' rsync
scripts to make sure the storm data needed to generate HWRF/GFDL guidance
reaches both the production and development machines simultaneously, a
capability that did not exist before.
5. Other Techincal issues: The printer in the HSU ops areaa did not work.
The Hurricane Specialists forced the printing of both TWOs from AWIPS. The
TWO products could not be seen in NWX (N-AWIPS) but they appeared on the
web correctly. Turned the printer off and on, which solved the problem.
1. Socket servers down. NWS CP (Office of Central Processing) began having
serious problems with multiple outbound circuits on Wednesday morning,
sometime around 1430 UTC. Several communication services were impacted,
including sockets, NIDS, and NHC's NOAA Port LDM feeds. As a result, some
products were ingested very slowly or were dropped entirely. It's possible
that surface obs were impacted as well. The most visible impact on NHC is
that text products were very slow to update on NHC's website. Some products
were stuck in a queue upstream of NHC. Some TAFB products had delay times
of 2 hours but this gradually improved to 15-30 minutes. The socket servers
were restored to full operational status at 1745 UTC.
2. A GOES-15 (GOES-West) satellite Yaw Flip maneuver, in which a satellite
is vertically inverted before/after eclipse season to extend its lifetime,
will begin at 2023 UTC on Thursday, September 22, 2016 and end at 0044 UTC
on Saturday, September 24.
For details, see:
GOES imagery scanning schedule:
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ncep.nhc.nco_contacts