Opened 7 years ago

Closed 7 years ago

#1273 closed help (fixed)


Reported by: wx019276 Owned by: um_support
Component: UM Model Keywords: header info, dump dates
Cc: Platform: MONSooN
UM Version: 8.4



I am trying to run an ensemble with the GC2 configuration, fully coupled. The way I am generating the ensemble is to use different atmospheric start dumps from around my experimental start date.

When I first tried to run with a start date specified in the umui as 2035/04/01 but a file that has header info as a start date 2035/04/02 the run failed straight away - and I was wondering if this was the problem. I thought that part of the reconfiguration dealt with this but maybe not?

I found ticket and used the change_dump_date mentioned there to alter the header info.

The concern I have is endiness as I am working on MONSooN. I brought the file back to Reading then ran change_dump_date on Jasmin which says that the file "is a byte swapped 64 bit ieee um file" I am just slightly concerned that the endiness is wrong.

I am currently doing a test it seems to be running - which I take as good sign but could it be that it is not doing what I think it is doing after this change if the endiness is wrong?

I was wondering if you knew if this would be a problem? Also is there is a version of change_dump_date on MONSooN that I can access? (I have looked but can't find one.) This would save me moving all the data back and forth just to change the header info.


Change History (6)

comment:1 Changed 7 years ago by jeff

Hi Ruth

The endianess of your data is fine, monsoon is a big-endian machine and the jasmin machines are little-endian, hence why jasmin reported the data as being byte swapped. All I/O in the UM is done in big-endian format regardless of the underlying endianess of the machine it is run on. If the file format was wrong your run would crash straight away as the input data would be nonsense.

You can find a copy of change_dump_date on monsoon here



comment:2 Changed 7 years ago by wx019276

Hi Jeff,

That is brilliant.

Thanks very much,

comment:3 Changed 7 years ago by wx019276

Hi Jeff,

I was also wondering if the ff2pp routine is available on MONSooN ( I previously have used it on hector) but I can't seem to see it on MONSooN. I can find it on ARCHER - can I just copy it onto MONSooN and then would it work there?


comment:4 Changed 7 years ago by jeff

Its not on monsoon at the moment, and you can't copy it from archer, I would need to build it for monsoon. But the archer executables should work on the postproc machine which can see your monsoon data. I have executables, including ff2pp, for the postproc machine under /home/jecole/linux_x86_64/bin.


comment:5 Changed 7 years ago by wx019276

That is geat

Many thanks again,

comment:6 Changed 7 years ago by jeff

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.