OT: cheap colo space in Southern Germany/Munich

[Eugen Leitl]

On Sat, Nov 24, 2012 at 07:20:19PM +0100, Philipp Kern wrote:

> The only thing I miss is an incentive to shift more onto IPv6. I think

The incentive can't achieve traction, since the end users don't
have access to IPv6, even if they knew what it was, which most
could hardly care less about.

> it's sane to let customers reconsider how many addresses they really
> need or move their business elsewhere. It's in line of address
> conservation. After all such a provider still wants to be able to
> accept new customers, too. On the other hand you already needed to
> justify your existing address space and growth beyond 20% more than
> what you currently have would be unjust to consider.

Whether the provider made a mistake, and decided to play by
the rules, while everybody else wasn't, it's not really 
changing the outcome. There's demand destruction, and
bigger fish eat the little ones, and then even bigger
fish eat the big fish, and so on.

Doesn't sound like much fun, so this is the time to GTFO pool.
 
> So the problem seems to be that it's a sudden price increase for
> existing customers, who relied on the prices not to change, even if
> their provider never promised such a thing. You are forced to push
> down the address conservation issue down to your customers and/or
> increase prices by, say, 5€, losing customers yourself. I'd say that

It destroys a whole class of business models. Including mine.

> all of this could've been anticipated given that the provider in
> question already went to monthly pricing of IPs for new customers.

You know, after they doubled the rack costs, yet promised to
keep me in the old contract I should have grown suspicious. 
Now that they're going to cheerfully screw that promise I 
know there is no way I can continue to do business with them.
Once bitten, twice shy.
 
> If you don't dare to even think of changing your business model
> slightly, changing providers does seem fair, doesn't it? Iff you can

No, my business model is dead. I'm back to mice and pumpkins,
aka 6 years worth of labor and money out of my own pocket,
all down the drain. The entry bar is now too hopelessly high
to bother, unless you have very, very deep pockets.
I will go back to life sciences R&D and reselling physical, 
tangible products in order to fund that hobby.

> get it more cheaply elsewhere, free market suggests to move. The thing

The free market suggests to get out of the pool, as it will soon
reach the boiling point.

> I agree with you, even though you didn't write it explicitly: It sucks
> that Hetzner only applies this to their dedicated server products and
> not to their VM and web space products. This obviously favors them in
> their services. I guess the lesson is not to host at a competitor
> then…

I think Hetzner knows just as well they're fucked. Eating your customers
before being eaten on your own buys them quite a few years more time.

My lesson from this is -- just for my own, very small needs -- is to
virtualize from the start, and fast-flux your hosts based on minimal 
price differential.

I think many players will starve before IPv6 coverage is quantitative.
By that time, hosting will be likely extremely large scale, and
almost completely automated.