Open yyan823 opened 5 years ago
How much free RAM is available for your SPAdes job?
How much free RAM is available for your SPAdes job?
I'm assembling on a server and have requested 3TB RAM for this SPAdes job.
This looks a bit suspicious. Does the problem reproduce if you'd rerun SPAdes via --continue option?
I've been getting this same error at the same point in the script (see below).
I'm using version 3.14.1 to assemble 739.4M total reads of 150PE Illumina RNA-seq data. I've gotten the same error running with both 1250GB and 1950GB RAM. Rerunning with the --continue option generated the same error in addition to a warning indicating "no saved checkpoint". See attached log and param files.
Did anyone find a solution to this?
1:18:01.173 22G / 23G INFO General (kmer_splitters.hpp : 364) Processed 4367340974 kmers
1:24:42.374 22G / 23G INFO General (kmer_splitters.hpp : 364) Processed 8624923124 kmers
1:27:42.631 22G / 23G INFO General (kmer_splitters.hpp : 369) Used 10509639109 kmers.
1:27:43.539 417M / 23G INFO General (kmer_index_builder.hpp : 120) Starting k-mer counting.
1:31:01.503 167M / 23G INFO General (kmer_index_builder.hpp : 127) K-mer counting done. There are 10537230085 kmers in total.
1:31:01.504 167M / 23G INFO General (kmer_index_builder.hpp : 133) Merging temporary buckets.
1:36:23.488 167M / 23G INFO K-mer Index Building (kmer_index_builder.hpp : 314) Building perfect hash indices
== Error == system call for: "['/opt/packages/spades/3.14.1/bin/spades-core', '/pylon5/mcz3b2p/slerch/exports_raw_reads/ribodep_367/trimmed_final/assemblies/ribo/rnaspades/K45/configs/config.info', '/pylon5/mcz3b2p/slerch/exports_raw_reads/ribodep_367/trimmed_final/assemblies/ribo/rnaspades/K45/configs/mda_mode.info', '/pylon5/mcz3b2p/slerch/exports_raw_reads/ribodep_367/trimmed_final/assemblies/ribo/rnaspades/K45/configs/rna_mode.info']" finished abnormally, OS return value: -11
@sluxlerch Are you sure this amount of free RAM is available for your SPAdes job?
@asl I allocated a total of 2000GB for the job, so 1950GB should should have been available. This is also consistent with the number of service units I was charged for the job.
Edit: To clarify this job was run on the Bridges supercomputer maintained by Pittsburgh Supercomputing Center, this requires the user to specify the total memory they want allocated for the job and this in addition to the job run time are used to deduct service units from the users account.
Hi developers,
I'm assembling an eukaryotic genome with SPAdes 3.13.1 and ran into err code: -11 after K21:
Could you please help with the issue? Here I attached params.txt and spades.log. Thank you!
spades.log params.txt