Re: [conduit] Missing grids from some 0.5 gfs files on CONDUIT

Carissa,

Looks like the 12 UTC run came in complete this morning, with no short files. I'll keep an eye on it over the next several runs and let you know if I notice any more incomplete files.

Thanks!

Pete


On 04/24/2015 10:10 AM, Carissa Klemmer - NOAA Federal wrote:

Pete,

We put a fix in the just the GFS directories in time for the 12z cycle today. Could you let us know if you have any issues after that comes out? If it is successful then we will apply the fix to all directories.

Thanks,

Carissa Klemmer
NCEP Central Operations
Production Management Branch Dataflow Team
301-683-3835

On Thu, Apr 23, 2015 at 6:35 PM, Pete Pokrandt <address@hidden> wrote:

I'm assuming the problem hasn't been fixed yet, but just for info, so far the same thing has happened with the 18 UTC 23 run for

F021, F030, F036 and F087 (data is currently in through F204)

Pete


On 04/23/2015 03:07 PM, Carissa Klemmer - NOAA Federal wrote:

Pete, 

Apologies for the issues. We have just determined that this is a disk read/write issue which is why you see it as random. We have engaged our IT folks to resolve this issue. We will let you know as soon as the problem is fixed. 

Carissa Klemmer
NCEP Central Operations
Production Management Branch Dataflow Team
301-683-3835

On Thu, Apr 23, 2015 at 3:29 PM, Pete Pokrandt <address@hidden> wrote:

All,

We have noticed a strange problem with the grids coming in for the 0.5 degree GFS files. The problem started with the 12 UTC run of the 2015/04/22 run, and several files (though not always the same ones) have been missing grids in each run since. The complete files contain 366 grids, the incomplete files always contain 327 and are always missing exactly the following grids:

5WAVH 5-Wave Geopotential Height [gpm]:500 mb
ABSV Absolute Vorticity [1/s]:100 mb
ABSV Absolute Vorticity [1/s]:975 mb
CLWMR Cloud Mixing Ratio [kg/kg]:150 mb
CLWMR Cloud Mixing Ratio [kg/kg]:550 mb
CLWMR Cloud Mixing Ratio [kg/kg]:800 mb
CSNOW Categorical Snow [-]:surface:
DLWRF Downward Long-Wave Rad. Flux [W/m^2]:surface
HGT Geopotential Height [gpm]:1000 mb
HGT Geopotential Height [gpm]:300 mb
HGT Geopotential Height [gpm]:50 mb
HGT Geopotential Height [gpm]:highest tropospheric freezing level
HGT Geopotential Height [gpm]:PV=2e-06 (Km^2/kg/s) surface
HGT Geopotential Height [gpm]:tropopause
HPBL Planetary Boundary Layer Height [m]:surface
O3MR Ozone Mixing Ratio [kg/kg]:200 mb
PRES Pressure [Pa]:low cloud top level:
RH Relative Humidity [%]:400 mb
TMAX Maximum Temperature [K]:2 m above ground:114-120 hour max fcst:
TMP Temperature [K]:10 mb
TMP Temperature [K]:350 mb
TMP Temperature [K]:650 mb
TMP Temperature [K]:900 mb
TMP Temperature [K]:max wind
TSOIL Soil Temperature    Validation to deprecate [K]:1-2 m below ground
UGRD U-Component of Wind [m/s]:1829 m above mean sea level
UGRD U-Component of Wind [m/s]:70 mb
VGRD V-Component of Wind [m/s]:20 mb
VGRD V-Component of Wind [m/s]:500 mb
VGRD V-Component of Wind [m/s]:750 mb
VGRD V-Component of Wind [m/s]:950 mb
V-GWD Meridional Flux of Gravity Wave Stress [N/m^2]:surface:
VVEL Vertical Velocity (Pressure) [Pa/s]:0.995 sigma level
VVEL Vertical Velocity (Pressure) [Pa/s]:450 mb
VVEL Vertical Velocity (Pressure) [Pa/s]:700 mb
VVEL Vertical Velocity (Pressure) [Pa/s]:925 mb


I don't think it's a latency issue for us, because our latencies for conduit have been < 30 seconds or so for at least the past 3 days.

One curious thing, the files are complete on motherlode.ucar.edu, but the files that we have as short, have a later modification time than the rest. For example, for the 12 UTC 23 run, we were missing grids for the F021, F102, F123 and F252 times. Here's a subset of the times on motherlode that shows these files:

gfs.t12z.pgrb2.0p50.f018                23-Apr-2015 09:35   67M  
gfs.t12z.pgrb2.0p50.f021                23-Apr-2015 09:49   67M  
gfs.t12z.pgrb2.0p50.f024                23-Apr-2015 09:37   66M  

gfs.t12z.pgrb2.0p50.f099                23-Apr-2015 10:00   66M  
gfs.t12z.pgrb2.0p50.f102                23-Apr-2015 10:42   67M  
gfs.t12z.pgrb2.0p50.f105                23-Apr-2015 10:03   66M  
gfs.t12z.pgrb2.0p50.f108                23-Apr-2015 10:03   66M  
gfs.t12z.pgrb2.0p50.f111                23-Apr-2015 10:04   66M  
gfs.t12z.pgrb2.0p50.f114                23-Apr-2015 10:05   66M  
gfs.t12z.pgrb2.0p50.f117                23-Apr-2015 10:06   66M  
gfs.t12z.pgrb2.0p50.f120                23-Apr-2015 10:07   66M  
gfs.t12z.pgrb2.0p50.f123                23-Apr-2015 10:51   66M  
gfs.t12z.pgrb2.0p50.f126                23-Apr-2015 10:09   67M  
gfs.t12z.pgrb2.0p50.f129                23-Apr-2015 10:10   66M  

gfs.t12z.pgrb2.0p50.f237                23-Apr-2015 10:43   66M  
gfs.t12z.pgrb2.0p50.f240                23-Apr-2015 10:44   67M  
gfs.t12z.pgrb2.0p50.f252                23-Apr-2015 11:32   62M  
gfs.t12z.pgrb2.0p50.f264                23-Apr-2015 10:49   62M  
gfs.t12z.pgrb2.0p50.f276                23-Apr-2015 10:50   63M  

We are feeding redundantly from conduit.ncep.noaa.gov and ncepldm4.woc.noaa.gov

Ideas?

Here is a (I think) complete list of the times that had short files since 12UTC 4/22. I'm guessing all of these files are complete on motherlode, but have a later modification time than the rest of the run.

2015042212_F081
2015042212_F112
2015042212_F141
2015042212_F156
2015042212_F159
2015042212_F195
2015042212_F228

2015042218_F063
2015042218_F081
2015042218_F198

2015042300_F006
2015042300_F018
2015042300_F021
2015042300_F087
2015042300_F114
2015042300_F120
2015042300_F138

2015042306_F006
2015042306_F069

2015042312_F021
2015042312_F102
2015042312_F123
2015042312_F252



Pete

-- 
Pete Pokrandt - Systems Programmer
UW-Madison Dept of Atmospheric and Oceanic Sciences
608-262-3086  - address@hidden

_______________________________________________
conduit mailing list
address@hidden
For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/

-- 
Pete Pokrandt - Systems Programmer
UW-Madison Dept of Atmospheric and Oceanic Sciences
608-262-3086  - address@hidden

-- 
Pete Pokrandt - Systems Programmer
UW-Madison Dept of Atmospheric and Oceanic Sciences
608-262-3086  - address@hidden

_______________________________________________
conduit mailing list
address@hidden
For list information or to unsubscribe, visit: 
http://www.unidata.ucar.edu/mailing_lists/