PGGB crash with no error

[SamCT]

Hello,

Thanks for the tool. I've successfully implemented PGGB on individual chromosomes just fine. I'm testing a new compute node, and PGGB hard crashed it. Having difficulty identifying the exact error, and am looking for feedback. First guess is that it ran out of memory (2Tb). The genome is relatively small (380 Mbp), and I have 30 haplotypes of high quality assemblies - repeat content is ~40%.

Command:

pggb -i CONCATENATE_ALL_RENAME.fa.gz -o out_pggb_hz_run_BORIS -n 30 -t 512 -p 90 -s 5k -m -S -D /my/hpc/TMP

Here is the last few lines of the log file:

[wfmash::align] alignment results saved in: /dev/stdout
wfmash -s 5000 -l 25000 -p 90 -n 1 -k 19 -H 0.001 -Y # -t 512 --tmp-base /my/hpc/TMP CONCATENATE_ALL_RENAME.fa.gz --lower-triangular --hg-filter-ani-diff 30 -i /my/hpc/TMP/CONCATENATE_ALL_RENAME.fa.gz.bf3285f.mappings.wfmash.paf --invert-filtering
11054547.45s user 89503.39s system 50039% cpu 22270.47s total 70891852Kb max memory
[seqwish::seqidx] 0.042 indexing sequences
[seqwish::seqidx] 82.059 index built
[seqwish::alignments] 82.059 processing alignments
[seqwish::alignments] 841.792 indexing

Curious to hear thoughts and suggestions to resolve this.

Thanks,
Sam

[AndreaGuarracino]

I would be surprised/sad if there is a memory problem with small genomes and 2TB of RAM! Can you add "\time -v" ad the beginning of the PGGB command? This way you should get the memory peak during the execution.

…

-t 512 is a record for me xD May be worth trying reducing it to a more polite 96/128, if it is really a memory issue due to just so many threads. Andrea

________________________________ From: Sam Talbot ***@***.***> Sent: Wednesday, August 20, 2025 17:31 To: pangenome/pggb ***@***.***> Cc: Subscribed ***@***.***> Subject: [pangenome/pggb] PGGB crash with no error (Issue #469) [https://avatars.githubusercontent.com/u/35304435?s=20&v=4]SamCT created an issue (pangenome/pggb#469)<#469> Hello, Thanks for the tool. I've successfully implemented PGGB on individual chromosomes just fine. I'm testing a new compute node, and PGGB hard crashed it. Having difficulty identifying the exact error, and am looking for feedback. The genome is relatively small (380 Mbp), and I have 30 haplotypes of high quality assemblies - repeat content is ~40%. Command: pggb -i CONCATENATE_ALL_RENAME.fa.gz -o out_pggb_hz_run_BORIS -n 30 -t 512 -p 90 -s 5k -m -S -D /my/hpc/TMP Here is the last few lines of the log file: [wfmash::align] alignment results saved in: /dev/stdout wfmash -s 5000 -l 25000 -p 90 -n 1 -k 19 -H 0.001 -Y # -t 512 --tmp-base /my/hpc/TMP CONCATENATE_ALL_RENAME.fa.gz --lower-triangular --hg-filter-ani-diff 30 -i /my/hpc/TMP/CONCATENATE_ALL_RENAME.fa.gz.bf3285f.mappings.wfmash.paf --invert-filtering 11054547.45s user 89503.39s system 50039% cpu 22270.47s total 70891852Kb max memory [seqwish::seqidx] 0.042 indexing sequences [seqwish::seqidx] 82.059 index built [seqwish::alignments] 82.059 processing alignments [seqwish::alignments] 841.792 indexing Curious to hear thoughts and suggestions to resolve this. Thanks, Sam — Reply to this email directly, view it on GitHub<#469>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AO26XHWHB4AZFOCA5DZNE5D3OSIGJAVCNFSM6AAAAACEL4GM5SVHI2DSMVQWIX3LMV43ASLTON2WKOZTGMZTQNJZGY4TIOA>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>