4

I have split a very large file(~500 MB) into many parts using the split command (Each split file has the same number of lines 131072, and of ~2.1 MB except the last one. Every line in each split file contains two integers only). The problem is with the split files. Some split files show size difference in nautilus and ls -l command output to great extent as shown for two files in the following image. enter image description here

What could be the reason for this? I am assuming ls -l output is correct for sure because files should be of same size.

Update:

When split files are copied to other drives they show comparable size for all the files(~2.1 MB as should be actually for the split -a 3 -l 100000 <filename> <file_prefix>). After opening the files from some other drive(after copying) and from the current drive , there is no difference in content(only the size displayed in both the drives is different). This problem is only observed while splitting large files using split. This problem occurs depending upon the way file is split(changing the number of lines in split command, sometime does not result in this problem). From above observation it looks like problem is not with split command but with the way nautilus calculates the size (May be depending upon the way files are stored or provided node points, but I don't have any idea how nautilus calculates the size). I am able to reproduce the same problem with 250 MB file for different split command setting. I will try to reproduce the problem for small file and if I can I will update.

It may be architecture dependent. I am using Ubuntu 16.04, Intel core-i5, RAM 16 GB, Hard drive ext4 format.

closed as off-topic by dessert, karel, David Foerster, Eric Carvalho, user117103 Nov 26 '17 at 19:45

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "Bug reports and problems specific to development version of Ubuntu should be reported on Launchpad so that developers can see, track and fix these issues." – dessert, karel, David Foerster, Eric Carvalho, guntbert
If this question can be reworded to fit the rules in the help center, please edit the question.

  • Can you share the files with us? – terdon Oct 11 '16 at 11:36
  • @terdon Where should I share the file?(can you name some preferable cite) Its 250 Mb file. Some more findings. If I copy the files to some other drive it shows the correct size but in same folder it does not. Also this phenomenon is dependent on the use of 'split' command i.e. in some cases this happens and in some it doesn't. – hunch Oct 11 '16 at 12:22
  • Why 250M? Your question only shows ~2.5M? Also, please edit your question and add the extra findings you mentioned above. Can you try to get a smaller file that reproduces the problem? You could try using head to get the first N lines of the file and see where the difference starts happening. – terdon Oct 11 '16 at 12:29
  • Unless you are using a locale I'm unaware of, I wouldn't trust a tool that display 2.1 MB as 20,64,874 bytes and 159.6 kB as 1,59,557 bytes. You'd rather show us the output of ls -l and wc applied to these files. – jlliagre Oct 11 '16 at 14:38
  • 3
    @jlliagre Output from 'wc' and 'ls -l' shows the same size, so the problem is with nautilus. This question can be closed now, mentioning this as the bug in nautilus. – hunch Oct 12 '16 at 11:55