I have a large file, approximately 8.5 GBs in size. Each time I gzip it the file and take a md5sum the checksum is different. If I gunzip the file and then take a md5sum of the gunzipped file it matches the original file's md5sum. So each gzipped file is valid yet has a different md5sum checksum. I.e. somehow gzipping the exact same file creates different, yet equally valid, gzip files. How can this be?
-I'm using the System Rescue CD 0.2.19 livecd (available here http://www.sysresccd.org/
-The version of gzip on the System Rescue CD is 1.3.3.
-I also got the same results with gzip 1.3.5 on Windows.
The commands I run look like this:
gzip -c bigfile.txt >bigfile1.txt.gz
gzip -c bigfile.txt >bigfile2.txt.gz
gunzip -c bigfile1.txt.gz >bigfile1.txt
gunzip -c bigfile2.txt.gz >bigfile2.txt
In this example the md5sums of bigfile.txt bigfile1.txt and bigfile2.txt should and do match. However, the md5sums of bigfile1.txt.gz and bigfile2.txt.gz do not match!? Why not? The resulting md5sums are:
As you can see from the md5sums the gunzipped image files are identical even though their gzipped counterparts are different. Why is this happening? When I perform test gzips on small files I can't reproduce this behavior. When working with my large gzipped files, however, this same thing happens every time. Can someone shed some light on this for me?