Hectic operation under WinServer 2019 / NVRAM / AMD Epyc config

Blaize

New member
Hi Benjamin,

we discussed about the temp file creation under windows server anno 2017 :) Then you helped me fixing it with the --bin feature. (Usually I run Diamond under virtual CentOS wo any issue, but in this case I need the raw power of the 64 cores / testing HW-SW)
So, sometimes Diamond stucks at the first computing alignment step with a 0 byte temp file, but if I restart the OS it is working perfectly creating a lot of neccessary tmp files. (Doing nothing with the bin option) Do you think that is there any relation again with this issue and the number of query bins?

Tested versions:
diamond version 0.9.36 / v2.0.0

Command:

diamond blastx -d nr -q sample_dedup.fa -o sample --threads 62 -f 100 -c1 -b18

Scoring parameters: (Matrix=BLOSUM62 Lambda=0.267 K=0.041 Penalties=11/1)
Temporary directory:
Opening the database... [0.067s]
#Target sequences to report alignments for: 25
Reference = nr.dmnd
Sequences = 284753542
Letters = 102465996361
Block size = 18000000000
Opening the input file... [0.027s]
Opening the output file... [0s]
Loading query sequences... [22.332s]
Masking queries... [5.682s]
Building query seed set... [0.124s]
Algorithm: Double-indexed
Building query histograms... [0.943s]
Allocating buffers... [0.008s]
Loading reference sequences... [43.412s]
Masking reference... [57.014s]
Initializing temporary storage... [0.005s]
Building reference histograms... [12.839s]
Allocating buffers... [0.217s]
Processing query block 0, reference block 0, shape 0, index chunk 0.
Building reference seed array... [58.729s]
Building query seed array... [2.295s]
Computing hash join... [113.553s]
Building seed filter... [0.193s]
Searching alignments... [35.566s]
Processing query block 0, reference block 0, shape 1, index chunk 0.
Building reference seed array... [13.585s]
Building query seed array... [0.451s]
Computing hash join... [102.443s]
Building seed filter... [0.167s]
Searching alignments... [27.205s]
Deallocating buffers... [31.723s]
Opening temporary output file... [0.001s]
Computing alignments... xxxxxxxxxxxxxxxx

Best regards & thanks:

Blaize
 

Benjamin Buchfink

Administrator
Staff member
Hi Blaize, no idea about this error unfortunately. Could you maybe run diamond using the --log option and show me the output, that might help me narrow down things a bit.
 
Top