1 |
On Wednesday 20 October 2004 2:37 pm, Jose Gonzalez Gomez wrote: |
2 |
> I also have this feeling when I read in the mailing list or forums |
3 |
> and see that a lot of people tell you to use tar/cp/rsync/whatever. |
4 |
> That's ok for your home machine, but I don't think those are solutions |
5 |
> for enterprise needs. Just imagine yourself telling your boss or client: |
6 |
> "Yes, I'm gonna have all those machines with your critical business data |
7 |
> backed up using rsync |
8 |
|
9 |
I backup mission-critical data with rsync on a nightly basis on a number of |
10 |
servers and workstations, and have never had a problem. If you're looking for |
11 |
tape backup solutions, you shouldn't be comparing apples to oranges. In the |
12 |
same way, I have a very hard time placing cp in the same league as rsync. |
13 |
|
14 |
I do agree that rsync has its limitations, however there are other similar |
15 |
solutions, such as rdiff, that can provide more robust functionality. |
16 |
|
17 |
|
18 |
> Joking apart, I think Gentoo needs a mind change to become a |
19 |
> corporate distribution and attract people doing real business... just |
20 |
> some thoughts of somebody who's trying to do Gentoo for a living and |
21 |
> would love to see Gentoo in every data center. |
22 |
|
23 |
"Real business" does not equal "support for tape backup solutions". While I |
24 |
can understand your comments in terms of practicality, I don't believe it's |
25 |
correct to widely generalize all fields of business, solely based on your own |
26 |
experiences with one corporation that has a specific set of requirements. |
27 |
|
28 |
There are two primary items of concern that I believe influence datacenter |
29 |
deployment of Gentoo the most: |
30 |
|
31 |
First, having hundreds of machines compiling updates on a nightly basis is a |
32 |
waste of system resources; this would necessitate some type of centralized |
33 |
package management system, which could in turn migrate the updated files to |
34 |
the machines without undue resource strain. The complexity of this solution |
35 |
increases proportionally to the variety of different system configurations |
36 |
that are being deployed. |
37 |
|
38 |
Second, in the case of having a large variety of servers that each provide |
39 |
different network services, keeping all of the configuration files up-to-date |
40 |
would be an enormously time consuming task without some type of automation. |
41 |
This is linked to the first concern, with a number of extra considerations. |
42 |
|
43 |
|
44 |
-- |
45 |
Anthony Gorecki |
46 |
Ectro-Linux Foundation |