lbcb-sci/ra

Ra dies and files are lost

Opened this issue · 1 comments

Hello,
I have been running ra on a ~5 Gb plant genome, highly heterozygous.
The input file is about 157 Gb of ONT reads (>QV7, above 10 kb), N50 24 kb. After about 5 days, the job dies while still at the minimap stage and the only message I have is:

[M::worker_pipeline::479382.631*42.28] mapped 30111 sequences
[M::worker_pipeline::479526.304*42.27] mapped 29912 sequences
[ERROR] failed to write the results

furthermore, the .paf file and the working dir are gone.
Is there a way to run separately (or even outside of ra) the alignment step?
I will test raven as well, but a colleague told me that it died as well (I don't know more details about this though).
Thanks

Hi Dario,
the minimap error indicates that you run out of disk space. You should certainly try raven for this as nothing is written to disk, but make sure to have at least 200 GB of RAM.

Best regards,
Robert