search

ablab / spades

SPAdes Genome Assembler
http://ablab.github.io/spades/
Other
752 stars 136 forks source link

Stuck at "Clearing raw paired index" #605

Open russellj7 opened 3 years ago

russellj7 commented 3 years ago

Hello,

I have been running metaSPAdes v. 3.14.1 on MacOS Mojave for Illumina HiSeq paired-end 2x150 data (about 25 million reads total).

Everything runs smoothly until it reaches the K55 stage and gets stuck at the "Clearing raw paired index" step. After sitting for a day, I killed it and tried restarting using the --continue option and it once again is stuck at this step.

Attached is the spades.log file. Any idea why it gets stuck and how to solve it?

Thanks!

spades.log

asl commented 3 years ago

@russellj7 This smells very strange and indicates some sort of corruption somewhere. Does changing the number of threads work?

russellj7 commented 3 years ago

@asl Thanks for your reply! Re-running this now and will post back when it finishes or if it gets stuck again. The computer it is running on has 36 total threads. I was running with 34 but bumped it down to 30 this time. Hopefully it works!

russellj7 commented 3 years ago

@asl I've re-tried the assembly noted above however with 30 threads instead of 34 but it again gets stuck on the "Clearing raw paired index" step. So instead I tried running the same reads through metaSPAdes (same version 3.14.1) on a Linux cluster environment and the job finished without error. So I'm not quite sure what went wrong on macOS. But anyway, everything worked out!

Out of curiosity, could there be any other possible reason this happened besides the number of threads?

Thanks!

asl commented 3 years ago

It smells like some bug around memory allocation on Mac OS, frankly speaking...