[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GNU Parallel Bug Reports inefficiency of --pipe -L
From: |
Ole Tange |
Subject: |
Re: GNU Parallel Bug Reports inefficiency of --pipe -L |
Date: |
Wed, 19 Sep 2012 13:52:21 +0200 |
On Sun, Sep 9, 2012 at 6:39 AM, Martin Frith <address@hidden> wrote:
> Hello:
>
> The NEWS file mentions that '-L n --pipe' is useful for fastq. I tried
> this, and it works, but extremely slowly. On the other hand, using '-N'
> instead of '-L' works and is much faster.
>
> I'm using GNU parallel 20120822. Here is how to reproduce the problem:
>
> seq 100000 > fake.fastq
>
> # This works and is fast:
> time parallel --pipe -N10000 cat < fake.fastq > out1
>
> # These work but are extremely slow:
> time parallel --pipe -L10000 cat < fake.fastq > out2a
> time parallel --pipe -L4 cat < fake.fastq > out2b
> time parallel --pipe -L4 --block 1M cat < fake.fastq > out2c
Confirmed. See http://savannah.gnu.org/bugs/?37325
/Ole