1 |
Mr. Rand(), |
2 |
|
3 |
I tend to prone the Diskless approach for the reasons you are mentionning. |
4 |
You can easily switch between roots with a network boot just by modifying |
5 |
your dhcp config and rebooting a node. This is really neat since you can have |
6 |
the dev environment on the actual cluster, test the new root with new libs |
7 |
and simply reboot some available nodes to test them. Furthermore, this |
8 |
approach opens the way to having multiple boot profiles with |
9 |
application-specific orientations (/me is thinking of the hellish deal of |
10 |
parallel Matlab with a polluted environment and the booting into a really |
11 |
optimized one for real MPI work ;)...) |
12 |
|
13 |
As for the "un-ncecessary attacker blahblahblah... Put your head behing the |
14 |
firewall. Beowulf nodes aren't meant to be publically available if they are |
15 |
to be efficient. nonetheless, departmental clusters (by night) could aslo be |
16 |
very possible with the diskless approach (even more so since you don't modify |
17 |
the current OS, which is most probably some horribly expensive Windows with |
18 |
the latest and greates Office suite which the deparment thinks is more |
19 |
important that the licences for Matlab...t'is not like we're trying to do |
20 |
some scientific work here eh!... (oops...dropped that one)... |
21 |
|
22 |
Hehe, in any case, I have had only great experiences with diskless nodes at |
23 |
the moment and really hope to see the Gentoo community take off on Clusterd |
24 |
(with or without disks ;) |
25 |
|
26 |
Le Mardi 11 Avril 2006 19:11, Dice R. Random a écrit : |
27 |
> Hi all, |
28 |
> |
29 |
> On 4/11/06, Hanni Ali <hanni.ali@×××××.com> wrote: |
30 |
> > Gentoo Cluster Handbook |
31 |
> > I. Sys Admin Requirements - What reader must know before moving forward |
32 |
> > (maybe describe clustering types here?) |
33 |
> > i) Networking options etc. |
34 |
> > ii) Disked vs Diskless |
35 |
> > iii) more... |
36 |
> |
37 |
> I'm particularly interested in what people are using for diskless |
38 |
> nodes and managing upgrades of system images across the cluster. I'm |
39 |
> envisioning a system where I have a build environment in which I can |
40 |
> upgrade system software and test functionality on a development |
41 |
> machine and then take a snapshot of the system and copy that up to a |
42 |
> NAS device so that the nodes can then boot it. It would be even |
43 |
> better if I could specify which packages I wanted (or rather, didn't |
44 |
> want) on the system images so that I could avoid having un-necessary |
45 |
> and potentially attacker-friendly packages such as gcc and portage on |
46 |
> the cluster nodes. |
47 |
> |
48 |
|
49 |
-- |
50 |
Eric Thibodeau |
51 |
Neural Bucket Solutions Inc. |
52 |
T. (514) 736-1436 |
53 |
C. (514) 710-0517 |
54 |
|
55 |
-- |
56 |
gentoo-cluster@g.o mailing list |