Submit Hint Search The Forums LinksStatsPollsHeadlinesRSS
14,000 hints and counting!


Click here to return to the 'Swap partition size' hint
The following comments are owned by whoever posted them. This site is not responsible for what they say.
Swap partition size
Authored by: jalbrecht2000 on Aug 09, '04 04:14:17PM

The reason I opted for UFS over HFS+ was overhead. There is no need for a journaling file system on your swap partition. Journaling your swap file system will only hinder a systems performance. If the system were to crash there is no reason to have the swap journaled for recovery.

This coupled with the lack of fragmentation on UFS and a little research swayed me to use UFS instead. However either way will work.

---
__________
Justin



[ Reply to This | # ]
Swap partition size
Authored by: Anonymous on Aug 11, '04 08:53:44AM

You can turn HFS journalling off, which I do. And the comment about fragmentation not being a problem on a swap partition is still relevant. If you're smart and put nothing else on that partition, it essentially gets cleaned out each time you reboot.

Further, 1 GB swap is foolishly, dangerously low for a MacOS X install. I have 1 GB of installed RAM, and I -regularly- get over 2GB of swapfiles dropped on my disk. My personal "high-water" mark for my usage is 4 GB of swap files, and so I made my partition 6 GB. It is always safer to be a little generous, since MacOS X behaves very poorly when it starts running out of swap space. In the days of 10.1, you could panic your machine. Panther seems to be a bit more tolerant, but odd errors will occur.



[ Reply to This | # ]
Swap partition size
Authored by: lerici on Jun 24, '07 07:42:19PM

I agree with Justin's statement.

UFS IS going to be SAFER for any partition where there are a lot of dynamic changes ... such as spool swap and tmp. Just for the reasons he gave.

In order to achieve "higher performance" hfs+ apparently gives up a great deal of stability. As as has been shown in some of the recent algorithms tomes the larger the tree the less it matters whether it is "organized" a certain way or not. In point of fact the more random it is "layered" the better. So as partitions grow in size becoming huge it may be that the performance gains achieved with hfs will become less and less.

Bottom line: Random good, complex trees may be bad.



[ Reply to This | # ]