Skip to contentSkip to conversation

inode_cache

DISCUSSION Swâmi Petaramesh 2 participants 2 messages Oct 14, 2012 – Oct 24, 2012 Btrfs file system

lore
Share & tools
lore
#0 lore raw

Hi there,

On a performance standpoint - and on Ubuntu Quantall kernel 3.5.0-17 - is it advisable to mount BTRFS with "inode_cache" ?

Is there any risk or counterpart doing so ?

TIA, kind regards.

Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E
Ne cherchez pas : Je ne suis pas sur Facebook.
#1 lore raw

Swami,
I believe the only thing that "inode_cache" mount option does is allowing to reuse inode numbers. Like if you had a file with inode number=X and you delete this file, then you will never have inode=X in the same file tree (subvolume) again. You can look at btrfs_find_free_ino() that allocates an inode number to understand this better. Since inode number is u64, it is perhaps unlikely that you ever run out of inode numbers. But still this mount option is present and implemented, so perhaps there is some reason for that. You can also look at "Free space cache writeback issue" note from Jan on the list; looks like he discovered some issue with that option.

Alex.

Quoted 17 lines · Swâmi Petaramesh

Hi there,

On a performance standpoint - and on Ubuntu Quantall kernel 3.5.0-17 - is it advisable to mount BTRFS with "inode_cache" ?

Is there any risk or counterpart doing so ?

TIA, kind regards.

--
Swāmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html