While Chris is attempting to fix his servers to restore the access to his GFS and ECMWF data I suggest everyone changes their settings to use Sam Bohler’s data and switch off ECMWF download and mixing.
Remember when after doing this you have to temporarily turn off WXSim auto run and change the Import function to use wdata.txt (or whatever name you use for the initial download) and then you can turn on auto run again.
Also remember WXSimLite (if in use) only accesses Sam’s database.
Once Chris has completed his rebuild and advises us accordingly then we can return to normal using his data.
Do you have any ideas how long the McMahon data will be unavailable? I have tried to remove all ECMWF downloads and set to Bohler but still getting errors about out of date data. We have been very pleased with the forecasts produced by WXSim and up to recently were using it as our main forecast, however with the lack of ECMWF in the mix there has definitely been a reduction in accuracy (Shetland, 120 miles north of mainland Scotland).
I think it’s great that Chris and Sam provide these data for no cost, however it does highlight how vulnerable the WXSim system is to a single point of failure.
My WXSim data has been fine as far as I can tell. Both GFS and ECMWF was available for all my forecasts today and no errors from WXSim. Last fo0recast at 15:00 today shows valid data for both as 06:00.
This is what I get when I run the forecast. WXSimate downloads without errors, the WXSim throws an error saying the GFS/ECMWF data are more than 24 hours old.
The data has been produced in time since the last big outage just over a week ago.
For anyone looking at the stats dashboard - I’ve reset the stats because there was some dodgy data from around the outage time that was messing them up. The stats will automatically rebuild as new runs are completed.
Hi Chris, I was getting error when I saw this post about the server being down, with the recommendation to change to Bohler until the server was fixed. Where is the dashboard?
