You're so right about those dangerous chemicals. They put dihydrogen monoxide in the ground. As you may know, dihydrogen monoxide kills thousands of people every year.
> Maybe your application, but not for others
Absolutely. As I mentioned, typical home use is one application where a quality SSD should be fine. I have a DNS server with several MBs of writes per day, so SSD would be fine for that (virtual) machine.
For the hot spares, no way SSD would work.
> the last server install I did with SSD's was 5 years ago, they're still going strong.
I hope you're checking those drives periodically and have some good monitoring. SSDs from five years ago had an expected lifespan of what, about five years in your application?
> who re-writes and entire HD 2x per day?
We would. Actually at that rate you'd expect it to die within 3-4 years. Drives dying is bad, so you need to replace them BEFORE they are likely to wear out. So figure you can write no more than 1/2 of the capacity per day.
For our hot spare server offering, we use raid arrays of 14 3TB drives, yielding 36 TB.
We'd like to use bcache or dm-cache. With a 500 GB SSD, we could write 250 GB / day - less than 1% of the array's capacity.
For typical home use, the write limit will allow five years or more of use. For other use cases, it's a deal breaker. Our servers are an example.
We offer a value priced combination hot spare server and backup solution. We'd like to use SSDs with bcache or something similar. We don't because we'd hit the write limit in three to six months. Write limits need to be 20 times higher before SSDs will work in our application.
> Some of that $100 goes back to supporting the connection to the customer, so carve out $30 of that $100 bill for that.
> This leaves $70/customer as revenue towards transit bandwidth. That is $7,000 in revenue for $500 in costs, so about $6,500 in gross profit.
That sounds all evil and stuff, so I'm sure it would go over great on potheadmoronhippie.com, but it's completely divorced from reality. Take as an example the most successful ISP in the United States, Comcast. Last year, they had $63 billion in revenue and $6.2 billion profit. That's under 10% profit. In the previous year, $4 billion profit on $56 billion in revenue - 7% profit.
The most successful ISP is making 7%-10% profit, which is slightly less than the 93% you tried to claim.
Thank you for that quote from the bill as it stands now. That's good that it includes "or if
98% of the time, judges are pretty good at seeing who is the asshole, who is the good guy, etc. Of course 2% of the time they screw it up and in those cases various web sites scream about it while significantly exaggerating the situation.
It's my understanding, and the GP who read the whole thing states, that the plaintiff pays the defendant's fees only if the sued IN BAD FAITH. The little guy can freely sue the big company if they have "a good faith belief" that the big company is infringing. It's pretty tough to prove bad faith, that the plaintiff didn't think they had a case. That comes into play when a plaintiff pulls crap like lying to the court about who their client is, and forging an inventor's signature - the crap the worst patent trolls do.
Once I found out that about 4-8 companies file half of all patent suits (and 90% of the troll ones), I figured it shouldn't be THAT hard to make it unprofitable for those companies to continue. Some say this bill isn't perfect, but if it manages to take enough profit our of trolling to stop those few big trolls, that largely solves the problem.
> yes i truly feel that cable companies go out of their way to help the consumer by lowering costs and operating at a minimal profit margin there by leaving no room for upgrades.
What the heck are you talking about? Did you read any of what I wrote, or did you just make up something to argue with?
> and never over subscribe.
That's precisely the point, that they DO over subscribe, and it would be wasteful for them not to subscribe enough that 2% of the time, it'll be oversubscribed in the sense that not everyone is getting the connection they'd like during those few minutes.
Picture a certain neighborhood. When people are at work 8:00-5:30, there are fewer than 200 people downloading at any given instant. From 6PM-9PM, no more than 300 people. After 9PM, again no more than 200 people. Each subscriber should get 10 Mbps. How much bandwidth does that neighborhood need? 300 people X 10 Mbps = 3,000 Mbps, right?
Right, that's what should be provisioned, 3,000 Mbps. Notice our usage pattern talks about 8:00-5:30 and 6:00-9:00. What about 5:30PM-6:00PM? As it happens, somewhere during that time period is when everyone gets home from work and checks their email, their Facebook, etc. For the busiest five minutes, maybe 5:40-5:45, there are 900 people downloading at once. Each wants 10 Mbps, so we need 9,000 Mbps, just for that five minute peak.
So, should we build 9,000 Mbps of infrastructure and buy 9,000 Mbps of backhaul in order to provide excellent service during those five minutes? If so, that money is being wasted 99.995% of the time. Spending three times as much, only to have that equipment sitting idle 99.99% of the time seems wasteful and stupid to me. The smart thing to do is buy about 3,000-5,000 Mbps of equipment and backhaul. That provides full speed 99.995% of the time. 0.005% of the time it's noticeably oversubscribed.
Note that it doesn't change anything if we upgrade and offer the customers higher speed. If we build 9,000 Mbps, we can provide the customer 30 Mbps
* I work on the server end of things. I don't know what the exact peak times are for residential service, but I do know it has high peaks, because I see part of those peaks hitting the web server. The web server serves multiple time zones, so it's peaks are spread, meaning the ISP sees higher peaks within each specific area.
You can do your own QoS within your home or office, but that does nothing for congestion on the ISP's network during peak periods. So yes, I'd also like the ISP do "do it for you", doing QoS on their network. Video, specifically, is bandwidth heavy and insensitive to latency, whereas VoIP, for example, is light on bandwidth and very sensitive to latency and jitter. So yes, I do want my ISP to handle my video and different from my VoIP. For my VoIP traffic, I want them to provide low latency and low throughput is fine. For video or FTP, I want high throughput and don't care about latency.
> When dealing with a faceless business it's important to use "power corrupts" and "Give them an inch
Agreed. Therefore, even though some prioritization would be good, I'd rather have none that have whatever monopoly ISPs think is good for their bottom line.
> Why can't you just pay more? Or maybe the ISP could upgrade their equipment to handle load.
This comment explains why that doesn't work:
Basically, to avoid any slowness, 2% of the time you need five times as much capacity as the other 98% of the time.
Why would you insult the meta-moderation script like that? How is a script supposed to know which articles are time-sensitive? Oh, were you under the impression that some human is involved in choosing stories, aside from us users who hit the meta-mod page once a month? I don't see any sign of intelligent life.
> really the monopolies need to upgrade the BW infrastructure but they won't
Upgrades are always nice. I'm glad they've upgraded from 28Kbps to 28,000Kbps in the time I've been using the internet. 100,000 Kbps would be even better.
However, impossible virtually impossible to upgrade so much that you never exceed capacity, and it would be really stupid to try . The thing is, in the busiest five minutes, users want 5 times the bandwidth that they demand during a normal "busy" period. In order to provide 10Mbps of clean bandwidth during that 5 minute period, you need infrastructure capable of delivering 50Mbps 95% of the time. You could either provide 50 Mbps all day but only bill for 10, which would be stupid, or artificially limit it to 10Mbps, letting 80% of your capacity go to waste, which would also be stupid. The only reasonable thing to do is build enough capacity to provide the full speed 98% of the time, and allow it to be slightly congested for a few minutes per day.
While I tend to agree with most people posting and I'm generally in favor of net neutrality, I also like playing devil's advocate, looking at both sides.
My SSH connection uses about 0.001 Mbps. Latency on SSH is really annoying, because it means each time you type on key you have to wait for that letter or number to show up on the screen. So for SSH you use very, very little bandwidth, but it needs to be low latency.
Netflix is opposite - it uses up 1,000 times more bandwidth, and latency doesn't matter at all (though jitter does). During peak hours, when the ISP is 1 Mbps short of perfect performance in a certain area, does it make more sense to annoy the shit out of 500 customers using SSH and other interactive low bandwidth applications, or should the one customer's Netflix packets get queued, which he won't even notice. (The Netflix movie will just begin one second later).
Given the very real choice of annoying 500 customers who aren't asking for much bandwidth vs. an imperceptible difference in one customer's movie, I think the choice is obvious. Better to not annoy any customers by giving the interactive packets priority.
That's what I'd want my ISP to do even if both connections are mine. I'd much rather have an unnoticeable 1% quality reduction in the YouTube video I'm watching than have lost or slow packets in my SSH. I WANT my ISP to discriminate between low priority, high bandwidth sites (video) versus high priority interactive.
It might also be useful to get real and talk about what this actually means in practice. YouTube and Netflix are HALF of the traffic load. Without those two, the existing infrastructure would deliver everything else TWICE as fast. Philosophical discussions are interesting, but at the end of the day, would you rather get stuff done much, much faster and allow the cat video to buffer for 1.5 seconds?
I would extend that to say don't ever tell the boss what they need to do in a way that implies they don't know how to do their own job. That can be tricky if you are recommending that they reverse their own decision. Don't "act like you're smarter than the boss".
What has worked for me and people working for me is to bring facts along with a "from a programmer's perspective this option looks attractive" recommendation. Change "programmer's perspective" to whatever is appropriate. For many years I did IT security. CxOs would sometimes ask "should we do this" or "what should we do". I try to remember to answer "that's a business decision that's up to you, but FROM A SECURITY PERSPECTIVE
The idea is to recognize and explicitly state that you are looking at it from a specialist's perspective, focusing mostly on one aspect of it. What you don't know, but the boss may know, as if they are planning on scrapping the entire project next month anyway. I can't tell the boss that we should upgrade X, because as far as I know the entire division that uses X may be getting laid off tomorrow. What I can tell the boss that that an upgrade to X would provide benefits Y and Z, at a cost of A.