exert from an email following the discovery that output030 was entirely missing:
Searching back through my notes, it (Oct-B) crashed on the morning of the 17th which looks like some hours before output031.
The error looked like a glitch at NCI. And the job just didn’t start. I simply restarted it, but the error then looked like it couldn’t read the previous restart because it didn’t finish writing the restart properly.
So I started the more stern approach of deleting the most recent restarts and starting from there.
Given that restarts from every 10th year are kept it is likely that I restarted from here (030).
Oct-B was up and running again later that day (also output031 etc are there).
It is curious that 030 musn’t have been there to begin with.
Here we rerun that year by re-starting from 020.
exert from an email following the discovery that output030 was entirely missing:
Searching back through my notes, it (Oct-B) crashed on the morning of the 17th which looks like some hours before output031.
The error looked like a glitch at NCI. And the job just didn’t start. I simply restarted it, but the error then looked like it couldn’t read the previous restart because it didn’t finish writing the restart properly.
So I started the more stern approach of deleting the most recent restarts and starting from there.
Given that restarts from every 10th year are kept it is likely that I restarted from here (030).
Oct-B was up and running again later that day (also output031 etc are there).
It is curious that 030 musn’t have been there to begin with.
Here we rerun that year by re-starting from 020.