1 |
thanks! I'll set it to 0% then. |
2 |
|
3 |
On Mon, May 10, 2010 at 12:48 PM, Willie Wong <wwong@××××××××××××××.edu>wrote: |
4 |
|
5 |
> On Sun, May 09, 2010 at 07:39:01PM -0300, Crístian Viana wrote: |
6 |
> > what exactly is this reserved block count? is it about the number of |
7 |
> inodes? |
8 |
> > does that mean that, by default, regular users can only use 95% of the |
9 |
> > inodes? and why did I use all these inodes? I don't think I have that |
10 |
> many |
11 |
> > small files on this partition... |
12 |
> |
13 |
> When the filesystem fills up, services can start failing left and |
14 |
> right because they cannot write logs, cannot write temp files, etc. At |
15 |
> this point human intervention is necessary: root has to log in and |
16 |
> clear out the disk. But if the $ROOT filesystem is completely full, |
17 |
> one may not even be able to log in and/or that one cannot do any sort |
18 |
> of maintenance that is needed. So you have some sort of circularity. |
19 |
> (In which case you have to reboot, perhaps using another medium...) |
20 |
> |
21 |
> The way out is to reserve some breathing room for root so that when |
22 |
> everybody else is having problems he can still get in and fix the |
23 |
> problem. |
24 |
> |
25 |
> The 5% is historical from days when disks are much smaller. If you |
26 |
> have a sensible partition scheme you only really need to reserve the |
27 |
> blocks on the $ROOT filesystem. If the partition in question (IIRC) is |
28 |
> only for /home, then you can just turn off the reserved blocks all |
29 |
> together. |
30 |
> |
31 |
> Cheers, |
32 |
> |
33 |
> W |
34 |
> -- |
35 |
> Willie W. Wong |
36 |
> wwong@××××××××××××××.edu |
37 |
> Data aequatione quotcunque fluentes quantitae involvente fluxiones invenire |
38 |
> et vice versa ~~~ I. Newton |
39 |
> |
40 |
> |