Well, running sambamba markdup results in an error when using --io-buffer-size=X with X being larger than 1024.
I read this in issue #169 and was not expecting it to be still an issue.
BAM: some BAMs, WGBS data, between 19GB and 35GB in size (created by bsmap)
sambamba: Version 0.8.0
params: --remove-duplicates --nthreads=24 --compression-level=4 --sort-buffer-size=8192 --io-buffer-size=X, X=1024 to 8192 in 1024 steps
Diskspace is plenty, RAM is 1TB, ulimit -n is 4096.
sklages
commented
2 years ago
Well, running
sambamba markdupresults in an error when using--io-buffer-size=Xwith X being larger than 1024. I read this in issue #169 and was not expecting it to be still an issue.bsmap)--remove-duplicates --nthreads=24 --compression-level=4 --sort-buffer-size=8192 --io-buffer-size=X, X=1024 to 8192 in 1024 stepsDiskspace is plenty, RAM is 1TB, ulimit -n is 4096.
Any idea what I missed?