Files related to my CS self study
./lfs.py -n 3
INITIAL file system contents:
[ 0 ] live checkpoint: 3 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 1 ] live [.,0] [..,0] -- -- -- -- -- --
[ 2 ] live type:dir size:1 refs:2 ptrs: 1 -- -- -- -- -- -- --
[ 3 ] live chunk(imap): 2 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
command?
command?
command?
FINAL file system contents:
[ 0 ] ? checkpoint: 14 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 1 ] ? [.,0] [..,0] -- -- -- -- -- --
[ 2 ] ? type:dir size:1 refs:2 ptrs: 1 -- -- -- -- -- -- --
[ 3 ] ? chunk(imap): 2 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 4 ] ? [.,0] [..,0] [ku3,1] -- -- -- -- --
[ 5 ] ? type:dir size:1 refs:2 ptrs: 4 -- -- -- -- -- -- --
[ 6 ] ? type:reg size:0 refs:1 ptrs: -- -- -- -- -- -- -- --
[ 7 ] ? chunk(imap): 5 6 -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 8 ] ? z0z0z0z0z0z0z0z0z0z0z0z0z0z0z0z0
[ 9 ] ? type:reg size:8 refs:1 ptrs: -- -- -- -- -- -- -- 8
[ 10 ] ? chunk(imap): 5 9 -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 11 ] ? [.,0] [..,0] [ku3,1] [qg9,2] -- -- -- --
[ 12 ] ? type:dir size:1 refs:2 ptrs: 11 -- -- -- -- -- -- --
[ 13 ] ? type:reg size:0 refs:1 ptrs: -- -- -- -- -- -- -- --
[ 14 ] ? chunk(imap): 12 9 13 -- -- -- -- -- -- -- -- -- -- -- -- --
Commands:
creat('/ku3')
open('/ku3'); write(); close('/ku3');
creat('/qg9');
Block 0,8,9,11,12,13,14 are active.
The -i flag definitely makes it easier to understand the commands.
INITIAL file system contents:
[ 0 ] live checkpoint: 3 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 1 ] live [.,0] [..,0] -- -- -- -- -- --
[ 2 ] live type:dir size:1 refs:2 ptrs: 1 -- -- -- -- -- -- --
[ 3 ] live chunk(imap): 2 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
create file /us7
write file /us7 offset=4 size=0
write file /us7 offset=7 size=7
[ 0 ] live checkpoint: 3 -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
---
[ 4 ] live [.,0] [..,0] [us7, 1]
[ 5 ] live type:dir size:1 refs:2 ptrs: 4 -- -- -- -- -- -- --
[ 6 ] type: reg, size: 0, refs: 0, ptrs: -- -- -- -- -- -- -- --
[ 7 ] chunk(imap) 5 6 -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 8 ] type:reg size:4 refs:1 ptrs: -- -- -- -- -- -- -- --
[ 9 ] chunk(imap): 5 8 -- -- -- -- -- -- -- -- -- -- -- -- -- --
[ 10 ] live i0i0i0i0i0i0i0i0i0i0i0i0i0i0i0i0
[ 11 ] live type:reg size:8 refs:1 ptrs: -- -- -- -- -- -- -- 10
[ 12 ] live chunk(imap): 5 11 -- -- -- -- -- -- -- -- -- -- -- -- -- --
One needs to look into the last imap to find out the active files and directories.
` ./lfs.py -o -L c,/foo:w,/foo,0,1:w,/foo,1,1:w,/foo,2,1:w,/foo,3,1`
Block 0,4,5,8,11,14,17,18,19 are live. Again the last imap helps.
Apart from producing more write operations, small writes produce produce more garbage.
./lfs.py -L c,/foo:w,/foo,0,1
./lfs.py -L c,/foo:w,/foo,7,1
The later writes with an offset, which makes the size in inode offset+write size.
With directory creation, data block has to be written immediately. With file, data block is allocated and written when the file is written to.
./lfs.py -L c,/foo:l,/foo,/bar:l,/foo,/goo -o -i
In both cases, a new entry is added to the parent directory. But with a link, no new inodes are required. The old inode has to be updated and thus copied though with the new ref count.
A random policy creates more fragmentation.