All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Is basement hosting a good idea?
I have a chance to get a static residential IP with 100Mbps connection in Southeast Asia. I have done some research, and it seems that older hardware (which I have access to) seem to be a nice fit. Something like Xeon E5 v2 / v3 (12c/24t ~ 28c/56t), and 1st gen Xeon D (16c/32t).
Would it be a good idea to basement-host a dedicated server and lease it out? Or, maybe make a NAT VPS service? I doubt I can get more than a handful of IPv4's, so renting out VPSes like OVH seems to be out of the question.
I can provide a UPS and save all VM state before shutting down, but any power outages will result in the server being down. I realize this is a downside for potential customers.
I also realize that if and when I actually proceed with this, I would need to get a provider tag to advertise any services here. And if I actually do this, it looks like it's going to be a hobby project, not big and professional like other providers here.
So... before pulling the trigger, what do you guys think?
Comments
The end is nigh.
If you have IPv6 and can beat WebHorizon pricing ($4/year for 256MB NAT container), I'll consider.
Ask @cociu
edit: more seriously, a residential connection can't compare with what can be found in decent datacenters. It's not just about the UPS and diesel generators in case there is a power outage.
Residential ISPs for the most part don't take very kindly to this kind of thing so bear in mind they might turn you off & tell you tough cookies.
Likely its concrete, so it would not burn that easy.
We had a few people that did basement hosting but kinda invulcated.
I don't know, some people even had the thought about putting a datacenter in a flat, non basement.
Turns out static and shit is a bitch.
I would not do this with a single 100Mbit uplink, if you can get Gigabit, symmetric, may be worth a try but no idea about power costs in Asia.
Is the basement prone to flooding ... at all?
Just put the rack on a pontoon.
It will surely float if flooded.
For a free hobby project? Yes.
For a for profit business, I would say no. Your customers don’t need the drama and neither do
You.
You're going to get a variety of responses from people telling you different things. Realistically, if you want to, then do it. If it makes sense to you, then just do it.
With that said. My opinion is pretty negative. I don't think it makes sense to do it and I don't think you should be given a provider tag for it. You're trying to keep expectations low ("basement hosting" and "it's going to be a hobby project") but once money trades hands I think it's a different ball game. Clients will have different expectations even if you tell them it's a hobby project. Even if it's extremely cheap, my opinion is that it provides no actual utility to a "customer" when they can pay a few dollars more for even a slightly better server in an actual datacenter.
I think you have a lot to gain here, with knowledge, experience, etc. But I don't think there's any value proposition for the customer. I don't think it's worth your time and money. I think you can probably spend the same amount of money in power, network, etc. to get an affordable colocation contract elsewhere and start with that.
My 2 cents.
If you tell people up front what it is and sell them access, we all get to laugh with you when they open threads complaining about 3 day outages. The only crime is concealing the facts.
They all started from the bottom, if you can do that why not?
It is true people mentioned a lot of reasons and advantages for clients to go for a reliable DC-hosted provider, but, I see one case where I would go for someone who offers basement hosting offer DC hosting:
A VPN for geo-locked services.
Most known VPN/servers providers will be detected when visiting geo-restriced services, so a residential IP might be great for this purpose. Also it might work for exotic locations too! Maybe locations in South America, Middle east or some regions of Asia that do not have stable hosting industry. For example UAE, Egypt, morocco, Peru...
Hello my friend. I think most of the people on this forum have had this dream at one point in time or another in their career, so it is understandable your thinking the same thing.
However, the biggest things to consider are the list cons to doing this that you overlook in your excitement and positive outlook that can really hurt you if you try this without doing your homework first:
Even with the best scrutiny of your customers there will always be abuse in one way or another, this could be e-mail spam, port scanning, torrenting, etc and you will be responsible for this and answering to the abuse. Most consumer ISPs are not very lenient about this type of abuse and you could quickly end up without a connection.
Most ISPs require you to purchase a business level product (costs between 5-10x more than consumer product) from them to resell their services and have policies against doing this on a consumer grade product -- so be hyper aware of the terms of service for your connection so that some weirdo who randomly downloads some shady content which brings cops visiting doesn't result in you losing your ability to have home internet, in some cases permanently.
You will need to be very specific when selling a service which is single homed and single power source, even with UPS's there could be a case where you lose power for an extended period and you have no way to do anything about this. Professional datacenters have multiple fail-overs for both network and power, this is what most are really paying for.
I could probably list many others, but my point really isn't the specifics but to make you also think of the long list of liabilities you take on doing such a thing. If you have the money, time and resources to invest and your long term goal is to build your own datacenter or rent space and you just need a place for the next 6 months to host some low traffic websites, then cool. If you think your basement is going to be a serious long term datacenter and you are not prepared to handle things as a business (as in you should register a business entity to do business with for liability, insurance, etc), you could end up personally responsible for your customers abuse/misuse of the service and with the right combination of events, with no internet, in jail, sued or a combination of all of the above.
my 2 cents.
Cheers!
Go for it.
For the love of the pie, go for it.
Apparently basement hosters are very much like crossfitters, vegans, Arch users, atheists, flat earthers and anti-baxxers when it comes to being vocal about their thing. Good to know.
vegans? how?
I consider vegans as evil clerics with anti-nature domain.
They hate plant matters so much that they are hell-bent on specifically murdering them non-stop.
Basement projects have got a really bad reputation here but I feel like it may be possible if some level of reliability is factored in for example, as you said maybe the ups. Uptime expectation obviously won't be 99.999% on residential connection necessarily and what if you go over their fair use policy. Add some back up for the storage. And keep it affordable, MAYBE you have a chance.
unless doing this for your hobby, business, friends and family, you in for some pain.
I’d say it’s not the best idea.
I have a server room at home (public IPv4 , IPv6, Mikrotik router + Fortigate 30E, 10Gbit flat fiber) and there are 3 servers there right now. Lastly I had major outage due to network issue - fiber has been broken during road renovation… 3 days on 25Mbit, because I have unlimited LTE as backup network, but this is unacceptable. Luckily not a production, I prefer to place production on Dedicated Servers 😁
In case of this outage on production, I’d be in a large trouble with customers…
I don’t think you have conditions to provide network and power redundancy, install UPS, Power generators… you also need to care about hardware , security and stuff like that.
HomeDC (This is how I call it) or basement hosting is fine as a hobby project… for you.
Not for commercial usage.
Absolutely agree with all of this, BUT @wpyoga I think you should go for it. Consider the pros and cons though.
If you do it, don't do it for the money. Rather, do it for the learning experience, as it's bound to fail. If it doesn't, congrats -- you can move into a real DC. Make sure people who are renting these NAT VPSes from you are doing it for hobby projects too (try to keep the customers to your friends only) and are fine with data loss and power outages.
If you do, I would route everything through a VPN.
Who dares wins.
100 Mbps, where? Indonesia?
What about attic hosting? If the house floods its less likely to go down.
Lol, I like your joke..
I wouldn't be surprised if it was Thailand tbf.
Problem is that 100 Mbit in South-East Asia usually means connectivity/access to the local market. International transit is terrible during peak time, especially for a residential client.
No joke. I assumed he is Indonesian by searching his username, Indonesian name.
Practically, what he think is doable. He can achieve that with cheap price, similar what on my mind some times ago.
Let's say, he got Telkom Indihome 100Mbps and got dynamic public IP (by default), he can do hosting on his home. Lot of people done this before. Here are some:
http://cbtsmkm2bjm.ip-dynamic.com
http://152721301145.ip-dynamic.com/
http://mandiritrans.ip-dynamic.com
100Mbps package cost Rp 935,000 (approx. $65) per month. So yes, his idea is profitable I assume.
But this provider Telkom Indihome, is rarely achieved 99% uptime in a month, so reliability is big concern. Also, you got shared port, so network sucks.