The following forum message was posted by christical at http://sourceforge.net/projects/s3tools/forums/forum/618865/topic/4664897:
I tried to look for a topic about this but couldn't find any. Sorry if it has
already been discussed.
I noticed something odd : depending on the use or not of --no-progress, I have
quite a different result in term of transfer speed.
For instance, the upload of a 2GB file :
# s3cmd put 2GBfile s3://test/
2GBfile -> s3://test/2GBfile [1 of 1]
2147479552 of 2147479552 100% in 220s 9.28 MB/s done
# s3cmd --no-progress put 2GBfile s3://test/
File '2GBfile' stored as 's3://test/2GBfile'; (2147479552 bytes in 28.9 seconds,
70.96 MB/s) [1 of 1]
Same thing with download :
# s3cmd get s3://test/2GBfiles3://test/2GBfile -> ./2GBfile [1 of 1]
2147479552 of 2147479552 100% in 221s 9.26 MB/s done
# s3cmd --no-progress get s3://test/2GBfile
File s3://test/2GBfile saved as './2GBfile' (2147479552 bytes in 19.1 seconds,
Those are the latest tests I ran but I got similar results everytime.
I'm using s3cmd with a Scality cloud (S3 like). Do any of you guys noticed this
Any idea how --progress (activated by default) works ?