1 |
On 04/06/2014 21:16, James wrote: |
2 |
> Hello everybody, |
3 |
> |
4 |
> I have a compliment of older x86 and amd64 boxes that I use for |
5 |
> testing things; new or specialized gentoo offerings and such. Mostly amd64. |
6 |
> Many of the older ide/ata drive systems have front loaded carriers |
7 |
> that have quick hot swap feature that make it easy to shut down |
8 |
> a machine. Swap the hardrive (and carrier cassette) and boot up |
9 |
> with a differnt OS for testing. Often I use one machine to work |
10 |
> on several differnet problems. I try to avoid the VM approach, |
11 |
> because much of what I work on involves actual hardware issues |
12 |
> and hardware verifications. |
13 |
> |
14 |
> I could put a minimal distro, such as system rescure, on to a usb stick, |
15 |
> CD/DVD for booting, mount the drive and dd over complete images |
16 |
> from various places, then a quick reboot to test a drive based distro. |
17 |
> |
18 |
> I have many differnt such gyrations ongoing and I need to deploy |
19 |
> a singular semanctic for boot a myriad of offering to test/code on. |
20 |
> |
21 |
> To just name a few: pentoo, lilblue, lxqt(4), lxqt(5), wrt, |
22 |
> embedded gentoo etc etc etc. It seems as though each is a "walk_once" |
23 |
> exercise and I'm loosing my mind. Throw on top of this, BTRFS, CEPTh |
24 |
> gluster, xfs, zfs I'm beginning to become too fragmented to stay focused on |
25 |
> the task(s) at hand. |
26 |
> |
27 |
> (Side rant) |
28 |
> If you read some of my (broken english) postings, they are mostly due to |
29 |
> hardware irritants that have the majoring of my limited brain cells |
30 |
> agitated to the point of illiteracy...... I use gmane (to post to |
31 |
> gentoo-user) and it use to have spell checking, or I hacked it, but, for |
32 |
> what ever reason, spellcheck in gmane seems to be gone now........ |
33 |
> |
34 |
> Anyway, I'd like to hear all of the ideas, including various disciplined |
35 |
> (structured) approaches I can take to minimize uniqueness in my lab and what |
36 |
> I'm currently doing (mostly). I also have usb sticks, but I've |
37 |
> found booting various offernings on usb, particularly older hardware, to |
38 |
> be too biosed_burdened. |
39 |
> |
40 |
> Bear in mind, I also have dozens of embedded boards, some x86, but |
41 |
> mostly arm based, that are also in the mix. As soon as some less expensive |
42 |
> arm64 (aarch64) boards become available, those too will become much more |
43 |
> prevalent in my lab. I need some new organizational (software and Image) |
44 |
> ideas. My hardware is very well organied on large, open racks with |
45 |
> lots of UPS power and easy physical access to each box/board. |
46 |
> |
47 |
> I'm also going to draw things up using (app-admin/rackview). |
48 |
> If/when I can take my network security to the next level, |
49 |
> I'd like to open up 1/2 of the machines to the gentoo community |
50 |
> for testing and debug on actual hardware. A portal to actuall hardware |
51 |
> resources. I basically need the ability to move around complete |
52 |
> system images between differment boxes. |
53 |
> |
54 |
> |
55 |
> thoughts and comments are most welcome! |
56 |
|
57 |
|
58 |
|
59 |
I have no idea how to improve your situation. but I do have a thought: |
60 |
|
61 |
You are not the only person to deal with this as there are many |
62 |
manufacturers of embedded items and it must be quite common to test many |
63 |
images with many hardware setups. |
64 |
|
65 |
So how do other vendors do it? If we look at their workflow, perhaps a |
66 |
useable method will filter up through the wetware :-) |
67 |
|
68 |
|
69 |
|
70 |
-- |
71 |
Alan McKinnon |
72 |
alan.mckinnon@×××××.com |