NS record question

Bob Vance bobvance at alumni.caltech.edu
Wed Mar 28 16:17:33 UTC 2001


Thanks, Jim.

>The point about release candidates is to provide sort of beta-test
>versions prior to the actual release. It's not unreasonable for them
>to shake out more bugs,

I think that we all know the point of the RCs -- and God knows to those
who are running the code they are very welcomed :)
Gimme all that you've got! :)

My point was that it doesn't help someone whose trying to decide when to
jump, for someone else to keep saying c'mon, it's OK -- and, oh, pay no
attention to the man behind the curtain.  And that man.  And that man
over there.


>Well fine. Stick with BIND8.

I will -- for the immediate, short term.


>BTW the "learning curve" with BIND9 is pretty small.
>It's trivial if your zone files are in good order.

At the risk of belaboring the point, my files are in good order,
legally, at least for BIND8.
I installed and ran BIND9.  I didn't see DDNS updates occurring from
DHCP.  I ran 'nsupdate' and got an error.  I tried 'rndc' and it didn't
work (and I knew there were gonna be some issues with 'rndc' from
monitoring the list).  I looked for the man page for 'rndc' -- is wasn't
there.  Sigh.  I had to spend time researching whether the man pages
would be installed by the release or not.  Then I had to manually
install the man page.  I saw something about controls and keys (which I
have basically just been cursorily monitoring on the list since I was
not going to BIND9 immediately) and didn't feel like reading and
learning about it at that moment.  Not a good start for a "trivial"
upgrade.

Now admittedly, I was taking the BIND8, new-release approach, just
hoping it might be easy and would work OK until I had a chance to look
further into the details.  I didn't even read the migration doco, which,
BTW, is *not* my typical procedure when doing something for "real" -- I
always try to do a lot of research and reading.  This is part of the
problem -- I just don't have the time to start digging into it right now
when I don't know how much there'll be.  (Although, I'm probably
spending more time writing about it than it would take to actually do
it! :)


As I said in another post,
    "For people who have already done so, it may seem trivial."

But if you are embarking on a trip to the store for a six pack (of Diet
Cokes :) and see traffic backed up for as far as you can see, you may
say,
    "Geesh.  What the hell is gonna be after this?
     I don't need those Cokes *that* bad.
     I think I'll just wait till this mess dies down  little.
    "
While someone else goes ahead and pushes through and it turns out to be
clear sailing after a right turn at the next block from then on.  Well,
yes for him it's easy to say it's trivial.

And it helps to hear you say that it is trivial !!
So as soon as I have time (in about 3 wks :), I'll dig into the BIND9
migr and other doco and have everything in good working order to my
satisfaction.



>for your definition of "stable". For some folks, BIND9 already is
>stable. Or stable enough.

One that doesn't have an RC every two weeks :)

But, all seriousness aside, I *did* say the "impression" of
non-stability, not actual stability :)


-------------------------------------------------
Tks        | <mailto:BVance at sbm.com>
BV         | <mailto:BobVance at alumni.caltech.edu>
Sr. Technical Consultant,  SBM, A Gates/Arrow Co.
Vox 770-623-3430           11455 Lakefield Dr.
Fax 770-623-3429           Duluth, GA 30097-1511
=================================================





-----Original Message-----
From: Jim Reid [mailto:jim at rfc1035.com]
Sent: Wednesday, March 28, 2001 9:53 AM
To: bobvance at alumni.caltech.edu
Cc: bind-users at isc.org
Subject: Re: NS record question


>>>>> "Bob" == Bob Vance <bobvance at alumni.caltech.edu> writes:

    Bob> 1) there is a learning curve on the processes and features.
    Bob> I know that it seems ridiculous, but my time is very much
    Bob> precious lately, and taking time to read the migration doco
    Bob> and figure out changes for starting and stopping and 'rndc'
    Bob> and keys, yada yada, is not something I want to do right now.

Well fine. Stick with BIND8. Lots of folk are still running BIND4 for
the same reasons. BTW the "learning curve" with BIND9 is pretty small.
It's trivial if your zone files are in good order.

    Bob> If I could just compile and go, like with new BIND8 releases,
    Bob> I'd be glad to install and run it to test it.

BIND9 was designed to slavishly follow the protocol specs. This means
that backwards compatibility with BIND8 for some things simply was not
an option: like illegal zone files.

    Bob> 2) we have the impression that BIND9 is not stable given the
    Bob> frequency of RC releases.

The point about release candidates is to provide sort of beta-test
versions prior to the actual release. It's not unreasonable for them
to shake out more bugs, especially on the more obscure OS platforms.
So what would you prefer, less frequent release candidates or more
prompt fixes to the bugs they uncover? And when a new release
candidate comes out, the notification lists the changes. If none of
them affect you or your platform(s), then you can just stick with the
release candidate that works for you. The next upgrade can be done
after the actual release is made.

    Bob> it just makes sense for me, personally, to wait until BIND9
    Bob> appears more stable.

... for your definition of "stable". For some folks, BIND9 already is
stable. Or stable enough.




More information about the bind-users mailing list