1 |
On Mar 10, 2012 10:09 PM, "Pandu Poluan" <pandu@××××××.info> wrote: |
2 |
> |
3 |
> |
4 |
> On Mar 10, 2012 8:33 PM, "Alex Schuster" <wonko@×××××××××.org> wrote: |
5 |
> > |
6 |
> > Hi there! |
7 |
> > |
8 |
> > Is there an advantage in putting the portage tree on an extra partition? |
9 |
> > |
10 |
> > Currently, I'm using reiserfs, because I read that it is efficient when |
11 |
> > using many small files. On the other hand I also heard that it tends to |
12 |
> > get slower with every emerge --sync. |
13 |
> > |
14 |
> > Space is no longer an argument in these days, at least for my desktop |
15 |
> > machine. But I would like to optimize for speed -- emerge -DputnVj |
16 |
> > @world takes quite a while to calculate, I assume this is because so |
17 |
many |
18 |
> > ebuild files have to be accessed. |
19 |
> > |
20 |
> > Any tips on this? Does it make sense to use a special file system just |
21 |
> > for the portage tree? What would be best? Would it help to re-create |
22 |
this |
23 |
> > file system from time to time in case it gets slower with every sync? Or |
24 |
> > wouldn't I notice a difference if I just used a big ext4 partition for |
25 |
> > all portage related stuff? |
26 |
> > |
27 |
> > Anyone using a compressed RAM file system for that? :) |
28 |
> > |
29 |
> |
30 |
> This had been my burning question when I was deploying the company's |
31 |
production server, and forced me to do some research: |
32 |
> |
33 |
> * reiserfs is amazingly fast for reads, but suffers on simultaneous writes |
34 |
> * reiserfs does not have inode limits |
35 |
> * reiserfs' notail affects performance greatly depending on the nature of |
36 |
the system: I/O-bound (use notail) or CPU-bound (don't use notail) |
37 |
> * reiserfs, if mounted without notail, is very space-efficient |
38 |
> |
39 |
> So, I end up with the following mix: |
40 |
> |
41 |
> * ext2 for /boot |
42 |
> * reiserfs for /usr/portage and /var/tmp (RAM is at premium; can't use |
43 |
tmpfs) |
44 |
> * ext4 for everything else |
45 |
> |
46 |
> This cocktail has been serving me well. I don't need advanced filesystems |
47 |
like ZFS, XFS, or btrfs, because my servers are virtualized, and the |
48 |
advanced features (e.g., snapshot) is handled by the underlying hypervisor |
49 |
(XenServer) and SAN Storage (we use NetApp). |
50 |
> |
51 |
> Rgds, |
52 |
|
53 |
Okay, I did a mixup: |
54 |
|
55 |
If the system is I/O-bound, *don't* use notail (saves on disk read/write). |
56 |
|
57 |
If the system is CPU-bound, *use* notail (saves on having to 'unpack' the |
58 |
tail from the metadata). |
59 |
|
60 |
In my situation, the bottleneck is the SAN Storage, so I don't use notail. |
61 |
|
62 |
Rgds, |