North/South halos too small for advection.

[ChrisWells]

Hi,

I've had a suite of UKESM (u-bg601) running for ~95 years, and it just failed on 21081001, with

[0] ????????????????????????????????????????????????????????????????????????????????
[0] ???!!!???!!!???!!!???!!!???!!!       ERROR        ???!!!???!!!???!!!???!!!???!!!
[0] ?  Error code: 15
[0] ?  Error from routine: LOCATE_HDPS
[0] ?  Error message: North/South halos too small for advection.
[0] ?        See the following URL for more information:
[0] ?        https://code.metoffice.gov.uk/trac/um/wiki/KnownUMFailurePoints
[0] ?  Error from processor: 218
[0] ?  Error number: 320
[0] ????????????????????????????????????????????????????????????????????????????????

I followed the advice on ​https://code.metoffice.gov.uk/trac/um/wiki/KnownUMFailurePoints and turned on Max wind output and High diagnostics.

But I'm unsure what to look at now? Looking in cylc-run/u-bg601 I can see the job.err file in /log , but can't see anything to go from there.

I can see in /share/data/History_Data that the model got to 20181201.

Do you know where I should look to see the max wind and other diagnostics? And how I could use that to fix my problem and continue this run?

Cheers,
Chris

[ChrisWells]

Hi,

just an update to this: I copied my suite to u-bh765, downloaded the restart files from my other suite for 21080101, and ran from there, and it seems to have worked (it's on 21091001 now) - but I am further confused by this. In hindsight, I shouldn't have expected this to work; I thought it would be bit-reproducible and therefore fail at the same point?

So my issue seems to be fixed, but I don't understand what's happened here; I'll leave this open as hopefully someone will be able to shed some light on this.

Cheers,
Chris

[grenville]

closed through inactivity