Running Cactus on AWS - Memory Issue

[tinaveit]

Hi @glennhickey! Following up with: DataBiosphere/toil#4773

I ran cactus with the following command-line for 2 species (following your suggestions from the previous thread), and it worked successfully:

cactus --consCores 32 --nodeTypes c4.8xlarge,r4.8xlarge --minNodes 0,0 --maxNodes 20,1 --nodeStorage 250 --provisioner aws --batchSystem mesos --metrics aws:us-east-1:xtremo-devo-cactus --defaultMemory 2000000000 --maxMemory 256000000000  --logFile cactus.log seqFile.txt output.hal

However, when I used the same command line to run 8 species, the cactus run crashed at the beginning and produced the attached cactus.log. It looks like a memory issue, but I am not quite sure how to solve it. Any help is appreciated!

Thanks!
cactus.log

[glennhickey]

You need to stick to released versions of cactus. This could be the AWS documentation being a little dated (I think it tells you to clone the master branch). Your workaround is either

• git checkout v2.8.0 (inside your current cactus directory)
• get cactus from https://github.com/ComparativeGenomicsToolkit/cactus/releases/download/v2.8.0/cactus-bin-v2.8.0.tar.gz