Slow performance when redirecting stdout to an existing disk file

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



While debugging a custom program that can write large output files to stdout, I noticed that the run time as displayed by the Bash 'time' prefix wildly varied from run to run. It turns out that the issue isn't just with my program. I can recreate it with 'cat':

$ time cat infile >outfile

If 'infile' is on the order of 140 MB, 'time' might show something as low as:

real	0m0.146s
user	0m0.000s
sys	0m0.109s
CPU %	74.29

or as high as:

real	0m0.328s
user	0m0.000s
sys	0m0.109s
CPU %	33.31

If 'outfile' doesn't exist, the 'cat' runs much more quickly:

real	0m0.082s
user	0m0.000s
sys	0m0.081s
CPU %	99.77

Likewise, if I arrange to 'rm' the file right before running 'cat', it runs consistently quickly:

$ time rm -f outfile

real	0m0.019s
user	0m0.000s
sys	0m0.019s
CPU %	99.93

$ time cat infile >outfile

real	0m0.081s
user	0m0.000s
sys	0m0.081s
CPU %	99.75

If I do these two commands in this order over and over, the timings for 'cat' are consistently fast.

In my custom C program, I have the same issue if I write to stdout or if I 'fopen' a named file like this:

fopen(outfile_name, "w")

If the file doesn't exist, the program runs consistently quickly:

real	0m0.191s
user	0m0.082s
sys	0m0.109s
CPU %	99.73

but if it exists the run times fluctuate quite a bit from run to run:

real	0m0.579s
user	0m0.072s
sys	0m0.147s
CPU %	37.85

real	0m0.473s
user	0m0.072s
sys	0m0.148s
CPU %	46.60

If I 'fopen' a named disk file right after doing unlink(outfile_name), I get consistently fast run times.

Note: I don't think this is an issue with buffered I/O to stdout per se. I say this because I'm seeing pretty consistent run times if I write to stdout _as long as the output file didn't previously exist_. Rather, the issue seems to be with _overwriting_ an existing disk file when I/O redirection is used.

Are there any known issues with the Linux kernel, 'bash', glibc, or whatever else might be involved with standard I/O that might cause this issue? If this is not a known issue, what's the best way to open a bug ticket?

I'm running Fedora 27 with the latest upgrades. The Linux kernel is 4.18.9-100.fc27.x86_64, glibc is 2.26-30, and bash is 4.4.23-1.

Thanks,
Dave
--
Dave Ulrick
d-ulrick@xxxxxxxxxxx
_______________________________________________
users mailing list -- users@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to users-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/users@xxxxxxxxxxxxxxxxxxxxxxx



[Index of Archives]     [Older Fedora Users]     [Fedora Announce]     [Fedora Package Announce]     [EPEL Announce]     [EPEL Devel]     [Fedora Magazine]     [Fedora Summer Coding]     [Fedora Laptop]     [Fedora Cloud]     [Fedora Advisory Board]     [Fedora Education]     [Fedora Security]     [Fedora Scitech]     [Fedora Robotics]     [Fedora Infrastructure]     [Fedora Websites]     [Anaconda Devel]     [Fedora Devel Java]     [Fedora Desktop]     [Fedora Fonts]     [Fedora Marketing]     [Fedora Management Tools]     [Fedora Mentors]     [Fedora Package Review]     [Fedora R Devel]     [Fedora PHP Devel]     [Kickstart]     [Fedora Music]     [Fedora Packaging]     [Fedora SELinux]     [Fedora Legal]     [Fedora Kernel]     [Fedora OCaml]     [Coolkey]     [Virtualization Tools]     [ET Management Tools]     [Yum Users]     [Yosemite News]     [Gnome Users]     [KDE Users]     [Fedora Art]     [Fedora Docs]     [Fedora Sparc]     [Libvirt Users]     [Fedora ARM]

  Powered by Linux