Running the following command:

for filename in ./seqtk_1/subsample_1/*_1.fq.gz;
do file=`echo $filename|sed 's/_1.fq.gz//'`;
filenopath=`basename $file`;
outputpath=./BWA/seqtk_1/subsample_1;
bwa mem -v 3 ./combine_reference.fa.gz ${file}_1.fq.gz ${file}_2.fq.gz | sambamba view -S -f bam /dev/stdin | sambamba sort -o /dev/stdin ${outputpath}/sorted_${filenopath}_BWA.bam;
done

I am getting an error message:

Sambamba-sort not enough data in stream

However, BWA continues to run.

Am I using the /dev/stdin wrong?

Thanks