- NANOG - lists.nanog.org

Linux, ECN and old firewalls
by Lee Watterworth 30 Apr '01

30 Apr '01

Hello all, Bumped into a problem where my firewall was refusing connections from a linux machine, found the reason and thought I would share: ============================== CONFIG_INET_ECN: Explicit Congestion Notification (ECN) allows routers to notify clients about network congestion, resulting in fewer dropped packets and increased network performance. This option adds ECN support to the Linux kernel, as well as a sysctl (/proc/sys/net/ipv4/tcp_ecn) which allows ECN support to be disabled at runtime. Note that, on the Internet, there are many broken firewalls which refuse connections from ECN-enabled machines, and it may be a while before these firewalls are fixed. Until then, to access a site behind such a firewall (some of which are major sites, at the time of this writing) you will have to disable this option, either by saying N now or by using the sysctl.

4 4

Re: California power ... unplugged.
by mike harrison 30 Apr '01

30 Apr '01

> regulation and management of the projects, the US programs were plauged by > competition for bids and shortcuts as well as more regional operators. Lowest bidder (and kickback) mentality affects networks as well as power production.... It's worse than doing it low buck and on a prayer.

1 0

Re: California power ... unplugged.
by Nathan Stratton 29 Apr '01

29 Apr '01

On Sun, 29 Apr 2001, Andrew Metcalf wrote: > However, Westinghouse was pretty successful with this in France, I think > almost all their plants are from the westinghouse program (something like 50 > plants).. Yep, they have sold them almost everywhere except the US. :-( -Nathan

1 0

RE: California power ... unplugged.
by Nathan Stratton 29 Apr '01

29 Apr '01

On Sun, 29 Apr 2001, Dave VanHorn wrote: > Dosen't have to be that way. > The French have a single design in service, all their plants are the same. > Any operator from any plant can operate any other, and they don't have to > design everything from scratch every time. We could learn from this. Actually Westinghouse tried to do this, they were going to get the licenses and everything and all you needed to do was get the operating license in your state. It was a good idea, but it was just 10 years to late. -Nathan

1 0

Contact at Applied Theory?
by Vivien M. 29 Apr '01

29 Apr '01

Hi. Can anyone point me in the direction of the email address of a clueful person at Applied Theory? I've emailed the address at http://puck.nether.net/netops/nocs.cgi nearly 24 hours ago, but gotten no response, and I'm getting slightly tired of any packets from here to EarthLink disappearing into some equivalent of Null0 or /dev/null somewhere on Applied Theory's network. TIA, Vivien -- Vivien M. vivienm(a)dyndns.org Assistant System Administrator Dynamic DNS Network Services http://www.dyndns.org/

1 0

RE: California power ... unplugged.
by Roeland Meyer 29 Apr '01

29 Apr '01

We don't have to ignore depleted nuke-fuel. For one thing, there are regen/breeder reactors. For another, fuel usage per year is much less than the environmental damage from fossils. Lastly, we only need the nukes until fusion plants are developed. They're working on NIF just down the street, at LLNL. The point is that, the enviros won't let us build fossils and the anti-nukes wont let us build nukes. NIMBYs wont let us build either one. Until that's settled, we wont get any new plants of either stripe. a.k.a. a self-destructive deadlock. I reiterate that it will be more than 5 years before we can get the first batch of new power plants online. The next sucking sound you hear will be businesses evacuating California, once they figure this out. "No tickee, no laundry" and no power, no business. By the time you figure out you need new plants, it's 4 years past starting to build one. Ironically, American companies are building and selling pocket nuke plants that are safe as houses ... for sale to North Korea and Russia. The $4.5B we just paid to the generator companies would have bought 5 or 6 of those. Maybe, the rest of the world deserves better than we do ... But, I digress ... you have been an excellent example of what I wrote about, thank you kindly. Anyway, I wont reply on this thread, on NANOG, anymore. It's getting too far O/T. It has been suggested that we find another list for those that want to discuss this further. I may post on that tomorrow. > From: Roger Marquis [mailto:marquis@roble.com] > Sent: Saturday, April 28, 2001 10:25 PM > > All true, but irrelevant to the people with cancer in the Ukraine > and elsewhere. Still, having worked in rnd.pge.com back when it > was a state of the art department, the consensus there was that > small nuclear plants were far safer than the large one's in vogue > before 3 mile island (whose core is now encased in concrete for > thousands of years). > > No question nuclear is clean but only if you carefully ignore > the danger of depleted uranium. But I digress, that's a problem > for future generations (if we're lucky). > > Roger > > > Sorry, but nukes are clean and safe. Sure people have died > from nukes, > > but millions have died from producing coal for plants. Why > do we build > > coal plants and not nukes? Because people don't care if > OTHERS die, if > > 100,000 people a year die from digging coal they are not in your > > community, that is better then the risk to THEM however small. > >

2 1

Young Network Operators (was RE: California power ... unplugged)
by Sean Donelan 29 Apr '01

29 Apr '01

On Sat, 28 April 2001, "James S. Smith" wrote: > Perhaps you just need some properly built nuclear plants. :) Ontario has New nuclear plant design has some interesting differences from the old plants. What caused these changes? Of course, there were several incidents which woke people up. But another change also occurred. Nuclear engineering fell out of favor among college students. There were fewer youngsters entering the industry after TMI, so the average age of nuclear engineers has increase. With age comes wisdom, or just senility? Older designs were "big" in every sense of the word. If you can only build one plant, you better make it the biggest plant you can. Old designs reflected a deep belief that you could maintain control. Control systems kept the system safe. The engineers were smart people, and had confidence in their ability to engineer their way out of trouble. Is this an error of youth? A young driver believes he can control the car in any situation, an older driver keeps the car out of situations requiring control beyond his ability. What does this have to do with network operators? With the downturn in the high-tech job market, and the apparent end of the 21-year old CEO getting $100 million in venture capital, will we see a similar increase in the average age of network engineers? And with the increase in age, will we see more conservative network operations coming out of companies? Or will we adopt the extreme slow conservatism of the telephone industry? Sometimes you need to accept some risk in order to innovate. If you never deploy something until it is perfect, it will take a very long time.

1 0

California power ... unplugged.
by Roeland Meyer 29 Apr '01

29 Apr '01

This is a follow-up of something I posted last November. The Enviro-Nazis are at it again, demonstrating why no new power plants have been built here, in the past 15 years. Hunter's Point: San Francisco proposed doubling capacity of the Hunter's Point facility by building a duplicate plant right next to the existing one. Local groups are protesting on environmental grounds. Coyote Valley: San Jose finally approved the construction of the CalPine power plant, needed to support the Santa Teresa, Coyote Valley, and Morgan Hill areas. But, local slow-growth advocates and enviro-nazis are protesting and threatening to sue San Jose. San Jose is asking the state to help smooth things over so that the plant can get built. Between the enviro-nazis and the no-nukes folks (they *are* different groups), California has been dead-locked on power plant construction for almost 20 years. The winning strategy; get locals to buy into the "not in my backyard" sentiment (an easy sell, it worked in stopping prison construction, as well). This, in a population so brainwashed, it actually wants to pay more taxes, even when they are already taxed at the 60% level. Further, this same population thinks that electric cars are the answer to the environmental problem, since they produce no pollution (a fallacy and contradiction of known facts, for those that don't know). In 20 years, CA has gone from a 300% over-capacity to the present shortage. In the past 8 months, energy costs have sky-rocketed 700-900% on only 3% increased usage. The reason; CA doesn't produce enough of its own power and, in the last year, has crossed over to being dependent on energy imports, rather than a supplier of energy. This is a crucial transition that CA was warned about over 25 years ago, when the no-nukes folks were very active here. Obviously, they didn't listen. Estimates are that, it takes two years to bring a fossil-fuel plant online and 3-4 years for a nuclear plant. However, fossil-fuel plants like all plants, have to meet the California Clean Air Act. This means that they may not be allowed to operate, due to environmental constraints, even if they were magically built tomorrow. Remember that California has its own, much stricter, environmental guidelines. These are guidelines that only nuclear plants can approach. Yet, this is in a state that is littered with "nuclear-free" zones. Bottom-line: 1) Expect the paralysis to continue for quite a while and then wait 3-4 years for plants to come online. Reasonable estimates are out past 5 years. 2) Expect increasing levels of power outages, both Summer and Winter (Stage 3, rolling blackouts). 3) Expect increasing energy bills, both in gasoline and electricity, on the order of 300-500%, in the next 5 years. 4) Expect a coming exodus of industry, to anywhere but California. 5) Expect ever more totalitarian regulations, to manage the long-term emergency (you'll have to get a license to plug a toaster into the power-grid). These regulations will not go away, even if we get enough power plants online. 6) Dick Cheney was correct, we need to start building nuclear plants out here. 7) Data centers and colo providers will come under increasing pressure to pay their "fair share" of power generation (this means that they get charged double or even triple). 8) Backup power generators will have to be replaced with units that can stand increased duty-cycles or even provide sustained long-term operation. 9) New colo facilities will have to face increased regulatory pressure and approval cycles. colo prices will go up, as a direct result of these increased costs. 10) Expect to see the return of the wind-fall profits tax, on a permanent basis. Caveats: 1) If the enviro-nazis get "nuked" tomorrow, this will only change the time table, by a year or two. Energy plants are not built by magic. 2) This entire scenario could reduce energy usage growth to about 1-2%. Usage will continue to increase. Thus, exacerbating the problem. 3) Houston could get "nuked", by the feds. -- MHSC is looking forward to doing business outside of California. :) -- ROELAND M.J. MEYER Managing Director Morgan Hill Software Company, Inc. TEL: +001 925 373 3954 FAX: +001 925 373 9781 http://www.mhsc.com mailto: rmeyer(a)mhsc.com

2 1

OT (possibly)
by Scott Walker 28 Apr '01

28 Apr '01

I'm looking for a good referance on using ipf on solaris 8. I have to setup a firewall and would like to use solaris. OTherwise it's off to linux for me.

2 1

RE: RIP and RIPv2, "The glue that makes the internet work"
by Roeland Meyer 28 Apr '01

28 Apr '01

That was last month's issue. I chuckled too. But, for a small end-point LAN, it's not bad. Consider it appropriate tech, applied in appropriate places. Even fully static routes aren't bad, on small enough networks. > -----Original Message----- > From: John M . Brown [mailto:jmbrown@ihighway.net] > Sent: Tuesday, April 24, 2001 6:51 PM > To: nanog(a)merit.edu > Subject: RIP and RIPv2, "The glue that makes the internet work" > > > > Latest Linux Mag has this really nice long article about how > RIP and its new version RIPv2 is the GLUE that makes the internet > work. > > I almost fell down on that. > > Oh, wait, I do know a couple of exchange points that wanted to or are > running RIP. No REALLY!! > > jmbrown > >

5 4