[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bug-tar] Tar vs later linux kernels
From: |
Gene Heskett |
Subject: |
Re: [Bug-tar] Tar vs later linux kernels |
Date: |
Sat, 07 Apr 2007 01:22:02 -0400 |
User-agent: |
KMail/1.9.6 |
On Saturday 07 April 2007, Paul Eggert wrote:
>Gene Heskett <address@hidden> writes:
>> It has now been 22 hours since my first post to you, with no reply as
>> yet,
>
>My goodness! Aren't we the impatient ones!
Sorry, I guess I'm getting spoiled by the instantainiousness of the
internet in my dotage. I can remember when bread was a nickle, gas was 8
cents & I rode a horse to school, all of which came to a screeching halt
on 7 Dec 1941 when I'd been 7 for about 2 months. Yeah, I'm an old fart,
born 20 years too late, or 20 years too soon, never figured it out which
would have been better. :-)
>Once you change the kernel, all bets are off, and it's safer to save
>everything when in doubt. If you happen to know that this particular
>change was safe then you can manually tell 'tar' not to back up that
>particular file; but you'll probably find it easier just to live with
>the duplicate output.
Its not a matter of one particular file, Paul. its a matter of tar
thinking that everything is new and must have a level 0 backup on this
pass. A level 0 on everything is around 45GB uncompressed, 33 or so
after gzip gets done with those that will compress. My vtapes partition
is 180GB, and is split into 21 vtapes, dumpcycle is 7, but could be
reduced to 5 if everything would settle down.
But following kernel development, more or less in the roll of the canary,
and running amanda suddenly have become mutually exclusive. That's not a
very pleasant way to spend my retirement years.
BTW, something I just saw after running 1.16.2 20070404 snapshot for one
pass only, it was reported by 1.15.1 that the time formats were invalid
in the --listed-incremental reference file for 2 or 3 of the tarballs
being generated by the subsequent pass using 1.15.1 from a fedora 6 rpm.
That is a bit worrysome also, in addition to the failure of the exclude
test (#12) when I ran a make test after building it.
Anyway, my one test run seemed to indicate that I was not on the right
trail of a fix, so I'll let TPTB decide what to do. Right now, the
kernel patch that did this change has been reverted. But it needs to be
done at some point, so there will be a bump in the road when its next
applied.
Thanks, I appreciate the time you took to reply.
--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
I can't complain, but sometimes I still do.
-- Joe Walsh