If you're only being hit a few dozen times per hour per machine then you're not talking about a very scary DDOS attack anyway. As you noted, that's only on the order or hundreds of requests per second even with tens of thousands of machines all attacking at once. It seems like a bad tradeoff to do intensive detection techniques like you were discussing on every single query to avoid serving traffic from a small DDOS attack should one arise.
Instead of trying to "detect" a DDOS, you can simply replace 1) with rate limiting. Say you're not willing to let any single IP address hit you more than 10 times a second, for example. You can even bucket rates. Say 10 is an attack.
For 2) instead of putting everything in a firewall you can have whatever is your closest-to-the-client server throw away a request, send an empty reply, send a captcha page or whatever else you think is appropriate given the rate being exceeded.
I also think you left out an important part 3). You need some way for your various entry points to talk to one another about the traffic they're seeing. Otherwise you're relying on the individual (in the rate limit method) or the DDOS network (in the detection method) to come in through the same path or you're risking them separately filling up each entrance before determining to block.
Load balancing is a given. Even to handle normal traffic you have to distribute it to your front ends somehow. But a sysadmin blacklisting IP address ranges? Oh my. You do not want this task to be done manually. A human, or a team of them, isn't going to be able to keep up with with a distributed attack, and they're going to make mistakes when determining which addresses to block. Remember that in a distributed attack they're not all coming from the same address and there's not much reason to expect them to come in some sufficiently narrow band for the idea of ranges to be useful.
Interesting. It sounds like it was the opposite for you what it was for me. It helped you to remember by forcing you to comprehend. It made it harder for me to remember by distracting me from trying to comprehend.
But that goes right back to my point. Just listening and paying attention is easier and more effective than trying to determine which points to write down and then noting them. It's not that you can't do that. You clearly can, and many people do. It's just that the effort doesn't always buy you much compared with just plain observation. It's nothing to do with transcribing everything you hear, which is note taking to such an extreme that I don't see anyone doing that unless an actual transcript is needed for some reason other than just comprehension.
I know it sounds like I'm being a smartass, but I don't mean to be. Really, I recommend just putting down the pen, closing the laptop, turning off the tablet, and just paying attention. Everybody is going to be a little different, but since you're asking for advice, that is mine. I found early on in high school that taking notes of the pen and paper variety takes away from attentiveness in favor of trying to become a stenographer. Effectively, my attention would be split between the process of note taking and the lecture itself. And an electronic device is just that plus even more distraction. I find that when listening if there's something I do truly need to review, I'm that much more aware of that need and can go look it up with another resource (the text book, a syllabus, proceedings, internet references) after the fact.
For in-home defense a shotgun is a really good choice. It's easy to use, which is important under pressure. And assuming you load shot rather than slugs, there's far less concern about a projectile penetrating a wall and doing unintended damage to something unseen than with something like a rifle or a handgun.
But she's specifically asking for a DIY solution, and is apparently pretty serious about it since she's already gotten a dog and a gun for security purposes.
The article discusses that. The news here is that the researchers think they have a modification of the design that does generate significant lift at subsonic speeds.
Why assume it's any different for the hunters than the fishers? They can also enjoy the hunting part and the eating part, and even the shooting part without enjoying the specific fact that the animal died.
Is that a paradox? It sounds like a special case of the holographic principle which says that any description of a volume of space can be encoded on the boundary of a region enclosing it.
There you go assuming that the labels are accurate in a practical sense. The joke is that California requires that warning on many so many chemicals with so tenuous a connection to cancer that it's basically impossible to use as an actual warning. That problem is exacerbated by the potential lawsuits when not issueing the warning, the fact that there's no exposure/penalty for warning unnecessarily, and the lack of specificity you noted. The net effect is that if you see such a notice you can rest assured that some chemical compound nearby that you may or may not actually be exposed to might possibly have some connection to cancer at some concentration that may or not actually be present... or someone just wants to cover his ass and not get sued. Not a lot of information content.
Your audience is programmers, so highly technical is not an issue. So what do intend to teach? How does Linux work, how is it organized, what is its structure? Gentoo stage 0. How do I use the new system you're making me use? Whatever all of your tools best supports, or if that's not a concern just go with the popular Ubuntu.
That is an argument against the particular business practices of Monsanto as well as with related legal structure and rulings, one with which I agree. That is not at all an argument against genetically modifying food plants.
Also, statements with which you disagree are not automatically "trolling" or "nonsensical."
1. It's bad because it's bad. That's not a "way" that's a restatement of the assertion. If you have a reason in mind, state it. 2. Absolutely. We must do everything in our power to make sure that the environment isn't contaminated with... genes. Did you know that genes have become so prevalent that some have even been found within the human body? 3. I am immune from this as all of my hats are made from 93% tin.
It's not true that the AMD lead was that short. The Athlon came out and was immediately on par with or better than Intel's Pentium IIIs. By the time it was thunderbird vs coppermine/tualatin the lead was pretty sizable. That lasted throughout the Athlon64/Pentium 4 period and into the Core's run until the Core 2 duos arrived. The gap was close for a while with Inte's multi-core processors generally superior, but as little as about a year and a half ago, AMD had the better offering in the X3 than Intel's Core i3. Competition is tight, which has been good for the rest of us.
Slashdot Mirror
If you're only being hit a few dozen times per hour per machine then you're not talking about a very scary DDOS attack anyway. As you noted, that's only on the order or hundreds of requests per second even with tens of thousands of machines all attacking at once. It seems like a bad tradeoff to do intensive detection techniques like you were discussing on every single query to avoid serving traffic from a small DDOS attack should one arise.
Definitely. For a mid size site, $400 a month is practically nothing. If you can make your DOS problems go away for that, do it.
Instead of trying to "detect" a DDOS, you can simply replace 1) with rate limiting. Say you're not willing to let any single IP address hit you more than 10 times a second, for example. You can even bucket rates. Say 10 is an attack.
For 2) instead of putting everything in a firewall you can have whatever is your closest-to-the-client server throw away a request, send an empty reply, send a captcha page or whatever else you think is appropriate given the rate being exceeded.
I also think you left out an important part 3). You need some way for your various entry points to talk to one another about the traffic they're seeing. Otherwise you're relying on the individual (in the rate limit method) or the DDOS network (in the detection method) to come in through the same path or you're risking them separately filling up each entrance before determining to block.
Load balancing is a given. Even to handle normal traffic you have to distribute it to your front ends somehow. But a sysadmin blacklisting IP address ranges? Oh my. You do not want this task to be done manually. A human, or a team of them, isn't going to be able to keep up with with a distributed attack, and they're going to make mistakes when determining which addresses to block. Remember that in a distributed attack they're not all coming from the same address and there's not much reason to expect them to come in some sufficiently narrow band for the idea of ranges to be useful.
Interesting. It sounds like it was the opposite for you what it was for me. It helped you to remember by forcing you to comprehend. It made it harder for me to remember by distracting me from trying to comprehend.
But that goes right back to my point. Just listening and paying attention is easier and more effective than trying to determine which points to write down and then noting them. It's not that you can't do that. You clearly can, and many people do. It's just that the effort doesn't always buy you much compared with just plain observation. It's nothing to do with transcribing everything you hear, which is note taking to such an extreme that I don't see anyone doing that unless an actual transcript is needed for some reason other than just comprehension.
I know it sounds like I'm being a smartass, but I don't mean to be. Really, I recommend just putting down the pen, closing the laptop, turning off the tablet, and just paying attention. Everybody is going to be a little different, but since you're asking for advice, that is mine. I found early on in high school that taking notes of the pen and paper variety takes away from attentiveness in favor of trying to become a stenographer. Effectively, my attention would be split between the process of note taking and the lecture itself. And an electronic device is just that plus even more distraction. I find that when listening if there's something I do truly need to review, I'm that much more aware of that need and can go look it up with another resource (the text book, a syllabus, proceedings, internet references) after the fact.
That's nothing at all like the prisoner's dilemma.
For in-home defense a shotgun is a really good choice. It's easy to use, which is important under pressure. And assuming you load shot rather than slugs, there's far less concern about a projectile penetrating a wall and doing unintended damage to something unseen than with something like a rifle or a handgun.
But she's specifically asking for a DIY solution, and is apparently pretty serious about it since she's already gotten a dog and a gun for security purposes.
Lowest common denominator or greatest common divisor?
I think MLA and APA styles are probably more common than Chicago. I'd bet not using a style guide is most prevalent.
The article discusses that. The news here is that the researchers think they have a modification of the design that does generate significant lift at subsonic speeds.
It got one of four for me. And the one it got was adblock which would be very easy to detect.
Why assume it's any different for the hunters than the fishers? They can also enjoy the hunting part and the eating part, and even the shooting part without enjoying the specific fact that the animal died.
It probably accelerates deuterium and maybe some tritium along with all the protium. So generically "positive hydrgen ion" is fine.
California already knew that.
Is that a paradox? It sounds like a special case of the holographic principle which says that any description of a volume of space can be encoded on the boundary of a region enclosing it.
There you go assuming that the labels are accurate in a practical sense. The joke is that California requires that warning on many so many chemicals with so tenuous a connection to cancer that it's basically impossible to use as an actual warning. That problem is exacerbated by the potential lawsuits when not issueing the warning, the fact that there's no exposure/penalty for warning unnecessarily, and the lack of specificity you noted. The net effect is that if you see such a notice you can rest assured that some chemical compound nearby that you may or may not actually be exposed to might possibly have some connection to cancer at some concentration that may or not actually be present... or someone just wants to cover his ass and not get sued. Not a lot of information content.
Your audience is programmers, so highly technical is not an issue. So what do intend to teach? How does Linux work, how is it organized, what is its structure? Gentoo stage 0. How do I use the new system you're making me use? Whatever all of your tools best supports, or if that's not a concern just go with the popular Ubuntu.
And with the added benefit of not being crushed to death by rubble in the process!
That is an argument against the particular business practices of Monsanto as well as with related legal structure and rulings, one with which I agree. That is not at all an argument against genetically modifying food plants.
Also, statements with which you disagree are not automatically "trolling" or "nonsensical."
Vitamin A, who needs that crap, anyway?
1. It's bad because it's bad. That's not a "way" that's a restatement of the assertion. If you have a reason in mind, state it.
2. Absolutely. We must do everything in our power to make sure that the environment isn't contaminated with... genes. Did you know that genes have become so prevalent that some have even been found within the human body?
3. I am immune from this as all of my hats are made from 93% tin.
It's not true that the AMD lead was that short. The Athlon came out and was immediately on par with or better than Intel's Pentium IIIs. By the time it was thunderbird vs coppermine/tualatin the lead was pretty sizable. That lasted throughout the Athlon64/Pentium 4 period and into the Core's run until the Core 2 duos arrived. The gap was close for a while with Inte's multi-core processors generally superior, but as little as about a year and a half ago, AMD had the better offering in the X3 than Intel's Core i3. Competition is tight, which has been good for the rest of us.