snakemake always report " MissingOutputException in line 44, Missing files after 5 seconds:

Question

I always get the same error report in my RNAs-seq pipeline by snakemake:

MissingOutputException in line 44 of /root/s/r/snakemake/my_rnaseq_data/Snakefile:
Missing files after 5 seconds:
03_align/wt2.bam
This might be due to filesystem latency. If that is the case, consider to increase the wait time with --latency-wait.

Here is my Snakefile:

SBT=["wt1","wt2","epcr1","epcr2"]

rule all:
    input:
        expand("02_clean/{nico}_1.paired.fq", nico=SBT),
        expand("02_clean/{nico}_2.paired.fq", nico=SBT),
        expand("03_align/{nico}.bam", nico=SBT)

rule trim:
    input:
        "01_raw/{nico}_1.fastq",
        "01_raw/{nico}_2.fastq"
    output:
        "02_clean/{nico}_1.paired.fq.gz",
        "02_clean/{nico}_1.unpaired.fq.gz",
        "02_clean/{nico}_2.paired.fq.gz",
        "02_clean/{nico}_2.unpaired.fq.gz",
    shell:
        "java -jar /software/Trimmomatic-0.36/trimmomatic-0.36.jar PE -threads 16 {input[0]} {input[1]} {output[0]} {output[1]} {output[2]} {output[3]} ILLUMINACLIP:/software/Trimmomatic-0.36/adapters/TruSeq3-PE-2.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36 &"

rule gzip:
    input:
        "02_clean/{nico}_1.paired.fq.gz",
        "02_clean/{nico}_2.paired.fq.gz"
    output:
        "02_clean/{nico}_1.paired.fq",
        "02_clean/{nico}_2.paired.fq"
    run:
        shell("gzip -d {input[0]} > {output[0]}")
        shell("gzip -d {input[1]} > {output[1]}")

rule map:
    input:
        "02_clean/{nico}_1.paired.fq",
        "02_clean/{nico}_2.paired.fq"
    output:
        "03_align/{nico}.sam"
    log:
        "logs/map/{nico}.log"
    threads: 40
    shell:
        "hisat2 -p 20 --dta -x /root/s/r/p/A_th/WT-Al_VS_WT-CK/index/tair10 -1 {input[0]} -2 {input[1]} -S {output} >{log} 2>&1 &"

rule sort2bam:
    input:
        "03_align/{nico}.sam"
    output:
        "03_align/{nico}.bam"
    threads:30
    shell:
        "samtools sort -@ 20 -m 20G -o {output} {input} &"

everything is fine until I add "rule sort2bam" part.

When I dry-run ,it goes fine. But when I execute it,it report error as the question describe. And Surprisely it run the task where it report it stuck in the background.But it always run the one task.like these:

rule sort2bam:
    input: 03_align/epcr1.sam
    output: 03_align/epcr1.bam
    jobid: 11
    wildcards: nico=epcr1

Waiting at most 5 seconds for missing files.
MissingOutputException in line 45 of /root/s/r/snakemake/my_rnaseq_data/Snakefile:
Missing files after 5 seconds:
03_align/epcr1.bam
This might be due to filesystem latency. If that is the case, consider to increase the wait time with --latency-wait.
Shutting down, this might take some time.
Exiting because a job execution failed. Look above for error message

[Sat Apr 27 06:10:22 2019]
rule sort2bam:
    input: 03_align/wt1.sam
    output: 03_align/wt1.bam
    jobid: 9
    wildcards: nico=wt1

Waiting at most 5 seconds for missing files.
MissingOutputException in line 45 of /root/s/r/snakemake/my_rnaseq_data/Snakefile:
Missing files after 5 seconds:
03_align/wt1.bam
This might be due to filesystem latency. If that is the case, consider to increase the wait time with --latency-wait.
Shutting down, this might take some time.
Exiting because a job execution failed. Look above for error message

[Sat Apr 27 06:23:13 2019]
rule sort2bam:
    input: 03_align/wt2.sam
    output: 03_align/wt2.bam
    jobid: 6
    wildcards: nico=wt2

Waiting at most 5 seconds for missing files.
MissingOutputException in line 44 of /root/s/r/snakemake/my_rnaseq_data/Snakefile:
Missing files after 5 seconds:
03_align/wt2.bam
This might be due to filesystem latency. If that is the case, consider to increase the wait time with --latency-wait.
Shutting down, this might take some time.
Exiting because a job execution failed. Look above for error message

I don't know what's wrong with my code? Any ideals? Thanks in advance!

Answer 1

As you figured out, & is the problem. Control operator & makes your command run in the background in a subshell, and this leads snakemake to think that job is complete when in fact it is not. In your case, its usage doesn't appear to be required.

From man bash on usage of & (stolen from this answer):

If a command is terminated by the control operator &, the shell executes the command in the background in a subshell. The shell does not wait for the command to finish, and the return status is 0.

Answer 2

I know how to solve, but I don't know why it works! Just delete the '&' in

samtools sort -@ 20 -m 20G -o {output} {input} &