On 01/30/18 14:10, Terry Barnaby wrote:
Thanks for the reply and trying. With your example its a bit different as you are
creating the tar and compressing. The compression will take quite a lot of CPU and
this is probably the bottleneck in your case.
No, it isn't
I used "tar -zcf lin.tar f27k/linux-4.14.15/" because you used "tar -xf
linux-4.14.15.tar.gz". So, your procedure was un-compressing the archive. So, I
was
thinking it would generate a similar workload.
Running tar -cf lin.tar f27k/linux-4.14.15/ produces similar results
[egreshko@meimei ~]$ time tar -cf lin.tar f27k/linux-4.14.15/
real 1m56.593s
user 0m0.762s
sys 0m8.914s
The above is reading from the server and writing on the client disk and doing no
compression
The below is reading the non-compressed tar on the client and writing to the server.
[egreshko@meimei f27k]$ time tar -xf ~/Downloads/linux-4.14.15.tar
real 1m52.371s
user 0m0.806s
sys 0m10.801s
So, either direction the times are pretty much the same.
If I un-tar directly on the servers disks in question (SATA hard
drives software
RAID-1) the untar takes 13 seconds rather than the NFS 3 minutes ...
Of course running it on the same system will be *much* faster.
--
A motto of mine is: When in doubt, try it out