[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [K12OSN] Memory and K12LTSP kernel ?

Brian Fahrlander <kilroy kamakiriad com> wrote:

Well not really: you load the old version, boot it, load the new
target kernel, boot it, then later throw away the old one.

Which is, of course, what RHN does if you allow it to do kernel upgrades.
The thing is, that an upgrade goes from one kernel version to another.
Hence, the old kernel has its modules in /lib/modules/2.4.xx-y.z while the
new one puts its modules in /lib/modules/2.4.aa-b.c, so they're not

I'm not talking about a vast difference in the kernels- just one
minor sub-rev will convince RPM/Apt/etc.  And anything need with the
kernels will be added at apt-time.

But in this situation, there's *no* difference in minor sub-rev. You're
going from a kernel and modules compiled *one* way to a new bigmem kernel,
both with the same version number. Hence the modules will get overwritten
under the running kernel, which is why I suggested the cautious approach.
Mind you, some people would just re-install . . .

...and...I've done it before.  :)

Since the proof of the pudding is in the eating, I'll defer. <g> Maybe I'm
seeing a problem where there isn't one. In any case, if the original poster
tries upgrading the kernel as it runs, and it works, great - and if it
doesn't, the repair technique is just what I described: boot the CD in
recovery mode, chroot, and install the bigmem kernel.

I can hardly wait to find out how he does it . . . <g>


--- Les Bell, RHCE, CISSP
[Who spent today grappling with installing Linux on an IBM mainframe -
count yerself lucky!]

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]