1 |
On Thursday, 25 February 2021 15:42:38 GMT Grant Taylor wrote: |
2 |
> On 2/25/21 2:51 AM, Michael wrote: |
3 |
|
4 |
> > A reinstall in this context is not a wholesale replace. |
5 |
> |
6 |
> ~blink~ |
7 |
> |
8 |
> > It implies obtaining the latest Stage 3 archive from a mirror, |
9 |
> > but retaining part of your current installation. Your /home, /etc, |
10 |
> > /var/lib/portage/world, plus any databases e.g. in /var/lib/mysql/ |
11 |
> > and your kernel config will be retained from your existing system and |
12 |
> > will not be replaced. Back these up first along with any particular |
13 |
> > customizations you have made, before you untar Stage 3, so you can |
14 |
> > restore them. |
15 |
> |
16 |
> Ah. You seem to be talking about what I would call an "in place |
17 |
> upgrade" for Windows. As in stalling n over top of n-1 or n-2. That's |
18 |
> definitely less disruptive than I was thinking. I was thinking that |
19 |
> fdisk and / or mkfs would be involved. |
20 |
|
21 |
Yes, it is an "in place upgrade", but bottom up. You upgrade the filesystem |
22 |
to the latest OS image, then drop in your own existing settings/configuration |
23 |
and finally run a system/kernel update. |
24 |
|
25 |
Unless you want to change your partitions you won't need to use fdisk. |
26 |
|
27 |
mkfs is advisable, it will clear out any old cruft and address any changes to |
28 |
the default system directories - e.g. where $PORTDIR, $DISTDIR may be in the |
29 |
latest Stage 3. |
30 |
|
31 |
A quick diff between your backup of make.conf and repos.conf against the new |
32 |
Stage 3 archive contents will inform you of changes to default settings. |
33 |
|
34 |
|
35 |
> > Then rsync portage, update all your @world packages and build a new |
36 |
> > kernel (make oldconfig). Spend some time merging existing application |
37 |
> > config files with etc-update to make them compatible with the latest |
38 |
> > versions of these packages, reboot and hopefully that should be all |
39 |
> > there is to it. |
40 |
> |
41 |
> I may end up /needing/ to go that route. For the moment, I'm going to |
42 |
> try the incremental updates. |
43 |
> |
44 |
> > Yes, it would have been, but what is the benefit of updating multiple |
45 |
> > packages many times over, instead of doing it just once? |
46 |
> |
47 |
> In some ways, this is a learning experience. As in it's a proof of concept. |
48 |
|
49 |
Yes, nowt wrong with that and a sound reason to try it out. In this case, |
50 |
time spent and any problem solving would be an investment in learning. |
51 |
|
52 |
|
53 |
> The computer in question spends 2/3 of it's life doing nothing but |
54 |
> idling a few programs. So, it spending time compiling and producing |
55 |
> heat is not a bad thing in this case. Especially when there's 10" of |
56 |
> snow on the ground. ;-) |
57 |
|
58 |
Neil highlighted the use case of a server/system which can't afford |
59 |
interruptions. I'd still reinstall as explained above, but do it offline, |
60 |
test, backup, and then untar over the live filesystem to keep downtime to a |
61 |
minimum. |