| 1 |
Daniel Iliev <danny@××××××××.com> posted |
| 2 |
20070506083444.279A29148B@×××××××××××××.com, excerpted below, on Sun, 06 |
| 3 |
May 2007 11:34:42 +0300: |
| 4 |
|
| 5 |
> In case you didn't know... |
| 6 |
> |
| 7 |
> There are several kernel configuration options you can tweak to make it |
| 8 |
> cache more aggressively. Also you could try XFS - it is known to be one |
| 9 |
> of the most hungry-for-RAM file systems. Please, have in mind that these |
| 10 |
> tweaks could be dangerous for your file system in case of power failure. |
| 11 |
> Consider using an UPS. |
| 12 |
|
| 13 |
I knew about these in general, but still good to post as others may not. |
| 14 |
|
| 15 |
You are talking write-caching here. I generally leave that pretty much |
| 16 |
alone, for the reasons you mention (corruption in case of kernel panic |
| 17 |
and/or power failure). That's also why I've chosen not to run XFS. (I |
| 18 |
run reiserfs and while I did have issues some years ago, early kernel |
| 19 |
2.4, I've had none since the introduction of data=ordered journaling and |
| 20 |
that as the default, even when I had faulty memory and was having fairly |
| 21 |
regular kernel panics as a result. I wouldn't have wanted to try that |
| 22 |
with big write caches and/or XFS!) |
| 23 |
|
| 24 |
The caching I had in mind was read caching. No risk there. =8^) |
| 25 |
|
| 26 |
-- |
| 27 |
Duncan - List replies preferred. No HTML msgs. |
| 28 |
"Every nonfree program has a lord, a master -- |
| 29 |
and if you use the program, he is your master." Richard Stallman |
| 30 |
|
| 31 |
-- |
| 32 |
gentoo-amd64@g.o mailing list |
Archives