Slashdot Mirror
Morality of Throttling a Local ISP?
An anonymous reader writes "I work for a small (400 customers) local cable ISP. For the company, the ISP is only a small side business, so my whole line of expertise lies in other areas, but since I know the most about Linux and networking I've been stuck into the role of part-time sysadmin. In examining our backbone and customer base I've found out that we are oversubscribed around 70:1 between our customers' bandwidth and our pipe. I've gone to the boss and showed him the bandwidth graphs of us sitting up against the limit for the better part of the day, and instead of purchasing more bandwidth, he has asked me to start implementing traffic shaping and packet inspection against P2P users and other types of large downloaders. Because this is in a certain limited market, the customers really only have the choice between my ISP and dial-up. I'm struggling with the desire to give the customers I'm administering the best experience, and the desire to do what my boss wants. In my situation, what would you do?"
>>>it doesn't sound at all like subby has the freedom to change the ToS or implement hard caps.
That depends. If the original contracts said "unlimited time" not unlimited gigabytes, then yes the ISP can move to a metered model. I'd implement relatively easy limits like "100 gigabytes maximum" with $1 for every gigabyte over the limit. This would catch the most egregious users, and any extra dollars can be used to add more lines to handle more people.
Oh and to justify it to the boss, I'd cite the recent court case which states ISPs may not discriminate against P2P traffic. i.e. It's effectively illegal to filter traffic, but not illegal to implement metered usage such that customers reduce usage voluntarily.
"I disapprove of what you say, but I will defend to the death your right to say it." - historian Evelyn Beatrice Hall
Yep - that's how they do it here in Australia and despite all the flak we cop on Slashdot about our metered ISP accounts, the user-pays system actually avoids a lot of the problems you see with ISPs overseas.
- P2P throttling? Not here.
- Artificial speed shaping or restrictions. Not here, unless you surpass your monthly limit on a flat rate plan.
- Forbidding servers on residential connections? Not here.
- Deep packet inspection and other traffic manipulation? Not here.
- Bad contention ratios. Not here (on the good ISPs at least).
The 70:1 contention ratio in the summary is pretty shocking ... good ISPs here (iiNet, Internode etc) have 10:1 or less and buy more bandwidth proactively, before they actually need it. They can afford to do that, and keep their links running at 50-70% capacity, BECAUSE it's a user pays system. Additional bandwidth use means more revenue for the ISP and hence it's attractive to them to keep their pipes un-congested and fast.
The other advantage is that light users can pay pretty small amounts for a basic connection. My parents just use email and so I put them on a TINY 1GB per month plan. They never even use more than half of that, and the cost savings are significant (consider that they pay only 20 bucks a month, but larger plans of 50, 100, 200 GB per month cost 60-100 bucks).
So if you absolutely cannot upgrade your links, the "bill, don't throttle" approach is more attractive. It's less work than setting up packet shaping infrastructure and rules, won't affect the large majority of your customers, and will make sure that top 5% of leechers keep their habit under control a bit better (or pay for a higher account, which means more money for you!).
Oh and one last thing. Don't bill for excess usage - just shape their connection. Because if Joe Sixpack gets a virus and their connection downloads 100s of GB without their knowledge, they are not going to want a huge bill. The way most ISPs do it in Australia is after you reach your monthly limit (say, 80 GB at 24 Mbps), they'll shape your traffic to a slower speed (e.g. 128 kbps). That's still fast enough to browse the web and stuff, but will ease backhaul congestion due to P2P etc.
That depends. If the original contracts said "unlimited time" not unlimited gigabytes, then yes the ISP can move to a metered model. I'd implement relatively easy limits like "100 gigabytes maximum" with $1 for every gigabyte over the limit.
This actually penalizes the guy who downloads a heck of a lot, but he times his downloads so they always run from 11 pm to 5 am.
While it rewards all those folks who download a 10th that, but always max out the link from 4:30pm to 9:00pm, with P2P, and streaming download, at the same time all the other subscribers are trying to surf the web and get decent performance.
Usage-based billing doesn't make any sense -- ISPs often get burstability pay for a CIR, to the 95th percentile.
Consumers should too... That is, you should be able to burst your connection to download files, for certain amounts of time.
Each subscriber should individually agree to how much bandwidth they get to use on a continuous basis, and how much, and how long they will be allowed to burst, before either being billed or capped.
It shouldn't cost you, unless you stay bursted (I.E. max out your connection all the time during peak hours)
And to be consumer friendly, they should provide better terms for off-peak hour time, to actually reduce the number of even normal downloaders.
In my opinion, the best solution is to strongly throttle large bandwidth usages (P2P, FTP and NNTP streams, etc) during the periods of near-capacity, and automatically relax the filtering during off hours.
That's one way... Here's another:
Instead of trying to choose which protocols are heaviest usage, traffic shape people based on what the actual criteria that you care about is: Too much overall usage over long periods.
In Linux terms, set up a HTB with a queue for every customer. Set the base rate to whatever your backbone speed is (1/70th of the customer's line rate), the ceil rate to their line rate, and give them a nice big bucket - say, 120 seconds times their line rate.
Then, people who are normal users - web surfing, downloading an occasional email attachment, etc - will go full bore, any time they want it. People who are bittorrenting will go full speed for a couple minutes, and then decrease down to whatever bandwidth is available. At night, if there's a lot of backbone free, it'll go fast. At 7 PM, they get best effort on whatever is available.
This is a very simplified example. You could additionally shape them so that their web and email will take priority over bittorrent when they're at the bottom of their token bucket, or other fine tuning...
The basic message I'd like to get across is: you don't have to shape based on protocol, because you care about the usage, not the protocol. Just shape based on usage, and let them work out which protocols they want to use.