Recently at work I had an issue come up where I needed to ftp a multi-gigabyte file up to a vendors site and they had mentioned that their ftp server had experienced problems with receiving such large files. So they recommended that I split the file up into smaller chunks.
Splitting
Here is an example of how I used the UNIX command split to chop the file up. For this example, NOTE: I’m going to use a file that’s ~347MB.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | # file to be uploaded % ls -l | grep exp_db -rw-r--r-- 1 oracle dba 347181617 Jun 25 15:06 exp_db.dmp.gz % split --help Usage: split [-l #] [-a #] [file [name]] split [-b #[k|m]] [-a #] [file [name]] split [-#] [-a #] [file [name]] # split file exp_db.dmp.gz % split -b 100m exp_db.dmp.gz segment_ ls -l | grep seg -rw-r--r-- 1 dbapps staff 104857600 Jun 30 10:14 segment_aa -rw-r--r-- 1 dbapps staff 104857600 Jun 30 10:14 segment_ab -rw-r--r-- 1 dbapps staff 104857600 Jun 30 10:15 segment_ac -rw-r--r-- 1 dbapps staff 32608817 Jun 30 10:15 segment_ad |
Here’s the basic form of the split command:
1 2 3 4 5 6 7 8 9 10 11 12 | split [ -b n | nk | nm] [-a suffixlength] [ file [name]] # -b switch, size of output files where: # n = bytes # nk = 1*1024 bytes # nm = 1*1048576 bytes # -a switch: the number of letters to use when creating a "chunks" suffix # (defaults to 2, e.g. aa, ab, ac, ... ) # file: the file you want to split # name: the prefix portion of the name of a "chunk" of the split up file |
Joining
And here is the command to join the segment_* files back into the original.
1 2 | # example join command % cat segment_* > exp_db.dmp.gz |
NOTE: For further details regarding my one-liner blog posts, check out my one-liner style guide primer.
June 30th, 2009, @ 8:12 pm |
Tags: 
Category: 
