1 |
Yiannis wrote: |
2 |
> While |
3 |
> googling a bit I read an article on ibm's site about linux containers |
4 |
> (LXC) which is supposed to finally land on the kernel. I think that this |
5 |
> might be worth trying as opposed to linux-vserver. |
6 |
> |
7 |
> |
8 |
|
9 |
|
10 |
I don't really know all the in's and out's of this argument, but I would |
11 |
desire to have vserver push to integrate stuff upstream, but the main |
12 |
developer seems happy with the status quo and has had many knock backs |
13 |
previously. As you point out, independently a bunch of people seem to |
14 |
be implementing substantially the same functionality, but without the |
15 |
prior history... Shame we can't avoid the duplication of work here... |
16 |
|
17 |
(One quite interesting patch included in the vserver kernel is a COW |
18 |
implementation of hardlink breaking. This is interesting for a class of |
19 |
problems such as rsync style backups, or obviously for any kind of |
20 |
duplicated shared pools of files. I would have thought this was an |
21 |
interesting feature to push upstream on it's own, but just to bring it |
22 |
to your attention in case it's useful for something else?) |
23 |
|
24 |
Anyway, vserver is also a fairly developed wrapper around the |
25 |
containers, so hopefully any new stuff will absorbed into that project |
26 |
and gradually it's patch will become smaller, but it really is a |
27 |
terrific solution to a whole class of problems |
28 |
|
29 |
Good luck |
30 |
|
31 |
Ed W |