Re: Upgrade Problem
On Fri 04 Jan 2019 at 13:41:33 (-0500), Felix Miata wrote:
> David Wright composed on 2019-01-04 10:19 (UTC-0600):
>
> > On Fri 04 Jan 2019 at 04:30:00 (-0500), Felix Miata wrote:
>
> >>> This partitioning scheme seems really odd and unwieldy.
>
> >> Indeed. Considering the absence of a sysadmin,
>
> > What's so unusual about that?
>
> Standing alone, absolutely nothing, but it wasn't standing alone....
(The OP is standing alone, leaving us aside.)
By snipping the rhetorical question that introduces my paragraph, it
now appears that "unusual" refers to the partitioning scheme. It
doesn't. It refers to the absence of a sysadmin. Here's the paragraph
again:
“What's so unusual about that? For a long time I ran linux and work but
didn't consider myself an "operator" or "sysadmin". Employees in those
categories ran closed shops of MS and Apple kit, plus a splinter group
running what they considered "proper" unix on kit that I couldn't
start to afford. Most of mine was 2nd hand PC cast offs.”
> >> absence of 2 possible primary partitions on sda,
>
> > If the OP partitioned an MBR disk intending to subdivide the
> > filesystem, then it might be expected that they create an extended
> > partition. Why bother with holding off until you've got two
> > primary partitions set up first?
>
> Off the top of my head:
>
> 1-trivial I know, but avoiding seeing fdisk report "Partition table entries are not in disk order"
>
> 2-less trivial: partitions not being in disk order
I don't understand. The time sequence would be
sda1=primary [ free ]
sda1=primary [ "sda2"=extended ]
sda1=primary [ sda5=logical free ]
sda1=primary [ sda5=logical sda6=logical free ]
sda1=primary [ sda5=logical sda6=logical sda7=logical free ]
sda1=primary [ sda5=logical sda6=logical sda7=logical sda8=logical possibly-free ]
What's out of order?
> 3-potential to have a primary partition added following a logical, thereby making following
> freespace unavailable for one or more added logicals (disappearing freespace).
With the scenario above, it would be usual to fill the disk with the
extended partition, so there's no possibility of adding another primary.
There may be occasional instances like the following squence, but
they'd be pretty rare in comparison:
sda1=primary sda2=primary sda3=primary sda4=primary
sda1=primary free sda3=primary sda4=primary
sda1=primary "sda2"=extended sda3=primary sda4=primary
> >> and the absence of sda6,
>
> > I assume that's swap.
Yes.
> Yes, but:
>
> # df -hl
> Filesystem Size Used Avail Use% Mounted on
> /dev/sda1 23G 23G 0 100% /
> /dev/sda2 ? extended (presumably)
> /dev/sda5 9.2G 6.0G 2.8G 69% /var
> /dev/sda6 ? ? ? swap
> /dev/sda7 1.9G 6.5M 1.7G 1% /tmp
> /dev/sda8 416G 103G 292G 27% /home
> ?freespace? ?
>
> Where is the logic responsible for the original allocations? Would any Debian Installer have done it
> without intervention from the admin? It looks like the work of a naive admin. Yet, OP claimed "I
> haven't messed around with partitioning since the early days of Slackware, and that was with a great
> deal of trepidation".
We have no idea without being told the earlier history of the scheme,
what the intent was at the time of creating them, and how their
subsequent use evolved.
Here's the partition table of this laptop. Care to guess it's
evolution?
Number Start (sector) End (sector) Size
1 2048 2050047 1000.0 MiB
2 2050048 2582527 260.0 MiB
3 2582528 4630527 1000.0 MiB
4 4630528 4892671 128.0 MiB
5 4892672 347348991 163.3 GiB
6 347348992 429268991 39.1 GiB /
7 429268992 511188991 39.1 GiB
8 511188992 883275775 177.4 GiB /home
9 883275776 883292159 8.0 MiB
10 883292160 892084223 4.2 GiB swap
11 892086272 892803071 350.0 MiB
12 892803072 894900223 1024.0 MiB
13 894900224 947329023 25.0 GiB
14 947329024 976773119 14.0 GiB
Constrained by an inability to repartition the disk, how would
you distribute a Debian system across it while wasting the
least space?
Cheers,
David.
Reply to: