Re: ipv6

Hello,

On 02 Mar 2014, at 16:29, Bob McConnell <[email protected]> wrote:

> Gil André wrote:
>> Noel:
>> On 01 Mar 2014, at 11:31, Noel Butler <[email protected]> wrote:
>>> Hi,
>>> But that is just "another" work around like is already in use, actually worse,
>>> since your IP lists would be wiped out on a system upgrade, not nice if its a
>>> host server with 1000 SSL domains
>>> 
>>> The workarounds are there, but none of them fix the "root cause" - slackware’s
>>> default networking's lack of capabilities, and its only going to get worse as
>>> time goes by with more and more RIR's and ISP's running out of ipv4 to issue.
>> Hmmm… What I don’t understand is that modifying the scripts to take IPv6 instead
>> of v4 does not seem all that hard. I must be missing something here.
>> — Gil.
> 
> What you are missing is the scope of the problem. It is not all that hard if you only have one or two machines to update. But it because a significant chore with even a handful of workstations, let alone a small server farm.

Well… I manage several hundreds of (CentOS/Red Hat) servers at work, so…

On the other hand, as I have said before, I manage these machines on IPv4.
Make of that what you will.

> On the other hand, for those of us that don't have IPv6 in our immediate future, shutting off that option is just as difficult. The last time I talked to the Time-Warner tech line, neither the first nor second level techs knew what IPv6 was, let alone when it would be available. So I normally blacklist all of those modules as soon as possible. With six computers in my den, that takes a lot more time that it should. I would like to see the startup scripts do that for me when the drivers notice there are no IPv6 routers available.

I am sorry, I still fail to see what is the problem. 

Noel’s problem, I can understand: servers need to come up quick, and if
updating a server farm means losing IPv6 configuration on all machines,
we are dealing with a major PITA here.

With only six home machines to manage, I recommend using some form of
system automation, such as Cluster SSH, Ansible or shmux. Unless I
don’t understand your problem at all?

(And, yes, I also have six machines to manage in my living-room)  ;-)

— Gil.