NERC is predicting California (and therefor Internet data centers
in the region) may be subject to almost daily rolling blackout
throughout the summer. Although most major Internet data centers
have backup generators, the historical reliability data everyone
uses is based on "normal" power conditions in the USA, not daily
rolling blackouts.
Is California really out of power? News reports indicate California
is consuming less power than the same time last year.
http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2001/05/03/MN202545.DTL
Energy consumption was down 9.2 percent, or 2,967 megawatts in March
compared with the same time last year. In February, the number was 8
percent, or 2,578 megawatts, and in January it was 6.2 percent or 2,091
megawatts.
So why are there power shortages in California?
http://www.latimes.com/business/reports/power/lat_suit010518.htm
A Times analysis of state data found that, throughout the last two months,
about 12,000 megawatts of production was offline, more than a third of the
peak power used in California on a typical day. That has been about evenly
divided between scheduled and sudden plant shutdowns.
By contrast, shutdowns in the same period of 1999 and 2000 took only
3,300 to 5,700 megawatts offline.
Why is 2 to 3 times more capacity offline this year compared with previous
years? I don't know. It appears the "supply" shortage is not due to
increased demand, or even the lack of new power plant capacity; but due
to the shutdown of existing capacity by generation companies at a higher
rate than normal.
Normally, when demand drops you would expect prices to fall. Consumption
is down between 6.2% and 9.2%. However, in California generating companies
have shut down power plants faster than demand fell, creating shortages
and prices have been rising.
It will be an interesting summer.
Sean Donelan <sean@donelan.com> writes:
Why is 2 to 3 times more capacity offline this year compared with previous
years? I don't know. It appears the "supply" shortage is not due to
increased demand, or even the lack of new power plant capacity; but due
to the shutdown of existing capacity by generation companies at a higher
rate than normal.
My understanding is that there is a seasonal "online / maint" cycle
for the generation plants, and that many of them were not taken
offline for normal PM several months ago because of energy usage
_then_... my unsubstantiated guess would be that just like letting
your car go 30,000 miles between a pair of oil changes,
procrastination of the maint cycle eventually comes home to roost and
that's what we're seeing now.
---Rob
There has been much talk of the introduction of new TLDs -- either new
ICANN/DoC ones or those of the likes of New.Net -- affecting the "stability"
of the Internet. And yet so far on all the lists, not just this one, I have
not seen a single example of how the "stability" would or could be affected
by such introductions. Can anyone give me even one example?
Tim
i don't see where there's a problem if there is a single root, but in the
cases where there are "add on" roots which a limited portion of the
internet can see, or even worse, competing roots which are advertising
competing versions of the same TLDs, the loss of symmetry has some
significant implications for support and operations.
for example, it will be possible to send mail to folks who cannot easily
reply because they cannot resolve your domain name, or vice versa. this
could dramatically affect support costs, especially at dial ISPs with
lots of naive users who when, confronted with this problem, will call the
support lines.
likewise, when support staffs are confronted with problems that involve folks
whose DNS they cannot see, or when they are seeing problems with one
version of my.net.fred and their DNS is showing them a different resolution
of my.net.fred, it's going to get messy and expensive. lower level staff will
not be competent to handle these issues in many NOCs, leading to more
escalations to higher tiers. clearance rates in call centers will decline.
how bad this is will depend entirely on to what extent alternative TLDs
penetrate in these scenarios. obviously this potential exists today, but
hasn't had a significant effect that i'm aware of, becaue the penetration
of the alternative TLDs is insignificant. if new.net or one of the others were
to take off to any significant degree, this would likely change. i decline to
guess what the increased support costs might amount to.
will it be the end of the world? no. but will costs (and thus rates) go
up? yes.
richard
Uh, this is a Forbidden Topic on this list. Even one example could
affect the stability of the Internet 
There has been much talk of the introduction of new TLDs -- either new
ICANN/DoC ones or those of the likes of New.Net -- affecting the "stability"
of the Internet. And yet so far on all the lists, not just this one, I have
not seen a single example of how the "stability" would or could be affected
by such introductions. Can anyone give me even one example?
it's not the stability. it's more the compatibility.
there's a large difference between the icann and new.net tld's
which are being introduced. to be blunt, everyone (okay, not
everyone) uses the same root servers to identitfy what zones
to serve. i'd be suprised to see if anyone had any issues with
the introduction (besides the heinous registration lottery) of the
icann accredited tlds, it's more of an issue with new.net when your
customer can't email his girlfriend at lolita@whore.house because
you're not serving dot house.
but please, if this thread continues, maybe let's stay on topic
of the operational sense (if there are stability issues) and not
relive the new.net threads of the past (compatibility).
Tim
-ken harris.
Hrmm, I don't know if this is a perfectly good example of how the
"stability" of the Internet is gonna be affected.
In my most humble opinion, I think it means that the addition of these
new >3 letters gTLDs will not be transparent to users. This is due to many
factors, among those are software writers. I believe that there are many
software packages that will fumble when they encounter >3 gTDLs, and some
have the com/net/org/k12/edu/mil/gov (and sometimes `int' too) hardcoded
as the "legal" gTLDs. This means that these software packages will
actually break, and will need to be updated, patched, or replaced, if the
software vendor is either no longer in existance, or for example, a
software version for your particular OS is no longer manufactured or
supported (there are many examples of that in the academic world for
instance).
my 2 cents,
--Ariel
There's an interesting sidebar in an old "America's Network" magazine
about why there was a shortage. It's too long to include here (and likely
too far off-topic), so here's a link.
They don't include this sidebar in their online edition, and I typed it in
after a similar debate over at the datacenter list:
http://www.inch.com/~spork/calpower.txt
Charles
Whether you agree or not, I'd of though that the issues of stability or
rather the problems created with setting up new roots was very obvious to
anyone of a technical background?
Example 1) Most of the Internet resolves one set of addresses, a subset
resolves a different set. Like dialling a phone number and it ringing a
different person depending on which phone company you use.
Steve
Ken
Yes, I was asking specifically about "stability" (with a request that
someone perhaps try to explain what that might mean other than a term to
scare people with) rather than "compatibility". BTW, I don't think ".house"
is a New.net TLD 
T
Most notions of "stability" are derived from a sense of steady state
reachability or connectedness. The non-ICANN gTLDs create perturbations
in portions of the net by changing the local semantics how to reach
certain destinations. While these perturbations do not spread (due to
policy), difference in the semantics of the perturbed network and the rest
of the net cause a lack of reachability, creating disconnected networks,
which destroy the "stability" of the net.
-Subhendu
Based on the email address, this was probably a troll.
We're Sorry, but you can no longer log in via the technologist.com hostname.
Please log in at http://www.iname.com from now on.
We regret any inconvenience this may have caused you. Thank you for stopping by.
Anyway, the fact that some of us now block BGP acceptance for the
new.net blocks, because it causes us support costs, would be an
argument that the Internet is less "stable".
"Tim Langdell, PhD" wrote:
Those packages are already broken. A little more won't hurt.
I mean, WTF do they do if you type perfectly legal domain names like
co.uk?
This is a bit like using IP addresses within private networks. We have
RFC1918 to specify some addresses which are specifically not routable on
public networks to minimize this confusion. That doesn't mean that some
people don't use unallocated (or even allocated) addresses within private
networks...
Perhaps we need a tld or a group of tld's which are analogous to RFC1918
addresses?
On Fri, May 18, 2001 at 08:04:14AM -0700, Tim Langdell, PhD exclaimed:
There has been much talk of the introduction of new TLDs -- either new
ICANN/DoC ones or those of the likes of New.Net -- affecting the "stability"
of the Internet. And yet so far on all the lists, not just this one, I have
not seen a single example of how the "stability" would or could be affected
by such introductions. Can anyone give me even one example?
can we PLEASE not start this holy war again? The last thread was more than
long enough to contain every perspective under the sun ... check the archives
at nanog.org if you need a refresher.
Perhaps we need a tld or a group of tld's which are analogous to
RFC1918 addresses?
This has been brought up a couple of times, but there are some pretty big
issues with it.
When somebody says "like RFC1918" you also need to include "problems with
RFC1918" in that scope. For example, private domain names allow for local
reuse of global identifiers that collide in nasty ways. What happens with
RFC1918 addresses when two orgs use the same global identifiers locally
and then need to interconnect: somebody has to renumber. The same is true
for .pri (or whatever) domain names, in that Cowboy Hats, Inc. and Cowboy
Boots, Inc. may both setup cowboy.pri domains, when they merge they have
to do a lot more work which means that any original savings (of which
there are none, if any) would have been lost.
Also like RFC1918, private domain names will leak out in unexpected ways
causing various problems. Cache poisoning was bad enough, it would become
horrific with overlapping domains.
There is some (as yet unpublished) research data that says ~20% of the
queries currently going to the root servers are for invalid TLDs (as setup
by .private internal operators). Endorsing the use of private domains will
make this much worse.
The best solution -- just like with addresses -- is to use real domains.
Advocating private domains causes more problems than it would solve.
Does that include blocking nameservers that respond authoritatively
for those TLD's? Oh wait, that probably doesn't matter since I saw
a certain company featured on FC yesterday...
William Allen Simpson wrote:
Why is 2 to 3 times more capacity offline this year compared with previous
years? I don't know. It appears the "supply" shortage is not due to
increased demand, or even the lack of new power plant capacity; but due
to the shutdown of existing capacity by generation companies at a higher
rate than normal.
I think more capacity is offline than previous years because of a large
number of factors. Some of the most important are:
1) The ISO has designated more "no touch" days than in previous years. This
makes it harder for plant operators to schedule their maintenance
effectively. Less effective normal maintenance means more emergency and
unplanned maintenance.
2) The PG&E bankrupcy combined with other utilities being unable (or
unwilling) to pay their bills has resulted in generators not getting paid.
Generators are hard pressed to keep their plants running when they aren't
being paid and don't know if they will ever be paid.
3) The increased prices earlier in the year prompted utilities to produce
more of their own power and to buy as little as possible from outside
sources. This means that more plants have used up their air pollution
allotments for the year. It also means that more maintenance is needed now
than would otherwise be required at this time of the year.
Personally, I'm fairly certain that most of the blame for our current
crisis rests with the courts who compelled the utilities to buy power
regardless of the price. There is absolutely no rational defense for this
ruling that I can imagine, and with no bargaining power, the utilities were
raped by the generators.
David Schwartz
And the addition of new TLDs would cause instability because ....? I am not
sure I understand your point. And also, who said anything about setting up
new roots? There are a host (sorry for pun) of ways of introducing new TLDs
without adding new roots. For instance the New.Net approach and the addition
of new ccTLDs and the addition of new gTLDs via ICANN/DoC (.biz, .info,
etc). The core question remains, why would the addition of a new TLD per se
(that is, in and of itself) lead to or be possibly likely to lead to
"instability" of the Internet?
TL
William,
I can assure you I am not a "troll". As most people know, I am the ex-CTO
(note, "ex") of New.Net. Come to think of it that might cause you to
classify me as a "troll" -- but that would be your choice, I guess.
So, an increase in support costs is "instability"? I am not sure I get that
...
Tim