1.2.0 - getting core dump on classify test

[matthewlinks]

I'm trying to check that my install is working by working through some of the examples in the readme.

When trying to get the classify example to work I'm running into a problem after the filters are built and it tries to start classifying.

User@server:~/readbouncer/29Aug2022/1.2.0/ReadBouncer-1.2.0-Linux/bin$ ./ReadBouncer --config ~/readbouncer/conf/classify_example.toml
IBF-build processed 2 sequences (8.09135 Mbp) in 0.194196 seconds (0.617932 Kseq/m, 2499.95 Mbp/m)
 - 2 sequences in 82 bins were written to the IBF
IBF-build processed 2 sequences (5.69078 Mbp) in 0.0590251 seconds (2.03303 Kseq/m, 5784.77 Mbp/m)
 - 2 sequences in 29 bins were written to the IBF
IBF-build processed 4 sequences (9.75088 Mbp) in 0.0994459 seconds (2.41337 Kseq/m, 5883.13 Mbp/m)
 - 4 sequences in 52 bins were written to the IBF
IBF-build processed 102 sequences (25.6866 Mbp) in 0.82183 seconds (7.4468 Kseq/m, 1875.32 Mbp/m)
 - 102 sequences in 318 bins were written to the IBF

Classification results of: /home/User/readbouncer/data/SampleZMCDataSet.fasta

free(): invalid size
Aborted (core dumped)

My toml

usage               = "classify"
output_directory    = '/home/User/readbouncer/data/classify_example/output'
log_directory       = '/home/User/readbouncer/data/classify_example/log'

[IBF]

kmer_size     = 13
fragment_size = 100000
threads       = 3
target_files  = ['/home/User/readbouncer/data/Zymo-Refs/Bacillus_subtilis_complete_genome.fasta', '/home/User/readbouncer/data/Zymo-Refs/Enterococcus_faecalis_complete_genome.fasta', '/home/User/readbouncer/data/Zymo-Refs/Escherichia_coli_complete_genome.fasta']
deplete_files = ['/home/User/readbouncer/data/Zymo-Refs/Saccharomyces_cerevisiae_draft_genome.fasta']
read_files    = ['/home/User/readbouncer/data/SampleZMCDataSet.fasta']
exp_seq_error_rate = 0.1

InterleavedBloomFilterLog.txt shows info messages about successfully storing IBF to file.

When I check output_directory the .ibf files are created and non-zero. The _classified.fasta and unclassified.fasta are 0 as per the issue.

configLog.toml ends with

[classify]
chunk_length = 250
max_chunks = 5
exp_seq_error_rate = 0.10000000000000001
fragment-size = 100000
threads = 3
kmer-size = 13
read_files = ["/home/User/readbouncer/data/SampleZMCDataSet.fasta"]
deplete_files = [
"""
/home/User/readbouncer/data/Zymo-Refs/Saccharomyces_cerevisia\
e_draft_genome.fasta\
""",
]
target_files = [
"""
/home/User/readbouncer/data/Zymo-Refs/Bacillus_subtilis_comple\
te_genome.fasta\
""",
"""
/home/User/readbouncer/data/Zymo-Refs/Enterococcus_faecalis_co\
mplete_genome.fasta\
""",
"""
/home/User/readbouncer/data/Zymo-Refs/Escherichia_coli_complet\
e_genome.fasta\
""",
]

Any guidance?

thanks

[JensUweUlrich]

Hi @matthewlinks

I'm actually looking into the code and try to figure out what the issue is. I could reproduce the error on Linux, but as far as I can see, it only occurs when using "classify". The "target" usage for adaptive sampling seems to be unaffected.

[merfre]

Hello @JensUweUlrich

Sorry for another message, I am just having the same error when doing a playback run despite updating to version 1.2.1:

sudo /home/merfre/ReadBouncer/build/ReadBouncer-1.2.1-Linux/bin/ReadBouncer --config ./test2_config.toml

709 bins were loaded in 0.122398 seconds from the IBF 28831 bins were loaded in 4.1305 seconds from the IBF Trying to connect to MinKNOW Host : 127.0.0.1 Port : 9502 Connecting to MinKNOW using the token: 7e068aa2-88e7-472c-b039-9c024443678b Connection successfully established! Waiting for device to start sequencing! Please start the sequencing run now! Sequencing has begun. Starting live signal processing! Connecting to Guppy basecall server on address : localhost:5555 Start receiving live signals thread Start basecalling thread Start read classification thread Start sending unblock messages thread free(): invalid size Aborted

I am using the same basecalling set up that we previously messaged about in #40 .

Thank you for all the help

[JensUweUlrich]

@merfre

It seems that you built ReadBouncer on your system. I guess you need to delete the whole build directory and repeat the entire build process. If you want to avoid that, you can also download pre-built binaries for Linux x64 and start immediately.

[merfre]

@JensUweUlrich

I tried that again (deleted the entire readbouncer directory and compiled using the instructions on the readme) and got the same error:

sudo /home/merfre/ReadBouncer/build/bin/ReadBouncer --config ./test2_config.toml

709 bins were loaded in 0.106683 seconds from the IBF 28831 bins were loaded in 4.89535 seconds from the IBF Trying to connect to MinKNOW Host : 127.0.0.1 Port : 9502 Connecting to MinKNOW using the token: b3050d0b-f996-4b92-abdc-bd4e1de3aa0f Connection successfully established! Waiting for device to start sequencing! Please start the sequencing run now! Sequencing has begun. Starting live signal processing! Connecting to Guppy basecall server on address : localhost:5555 Start receiving live signals thread Start basecalling thread Start read classification thread Start sending unblock messages thread munmap_chunk(): invalid pointer Aborted

I then tried with the pre-built binaries for my system and had the same error occur:

sudo /home/merfre/Downloads/ReadBouncer-1.2.1-Linux_x64/ReadBouncer-1.2.1-Linux/bin/ReadBouncer --config ./test2_config.toml

709 bins were loaded in 0.123347 seconds from the IBF 28831 bins were loaded in 4.58021 seconds from the IBF Trying to connect to MinKNOW Host : 127.0.0.1 Port : 9502 Connecting to MinKNOW using the token: b3050d0b-f996-4b92-abdc-bd4e1de3aa0f Connection successfully established! Waiting for device to start sequencing! Please start the sequencing run now! Sequencing has begun. Starting live signal processing! Connecting to Guppy basecall server on address : localhost:5555 Start receiving live signals thread Start basecalling thread Start read classification thread Start sending unblock messages thread munmap_chunk(): invalid pointer Aborted

[JensUweUlrich]

That's strange. Could you delete and rebuild the IBF index files with version 1.2.1. Maybe the files have been corrupted while saving using the version 1.2.0.