Best Way to Start a Website Hosting Service?

Kwirl writes "Lets say that I wanted to start a small business endeavor, namely reselling my server space and offering pre-built websites. What resources would I need to start something like this on my own? What hosting service would best suit those needs? What would be the best way to manage a subdomain-level service that provided a basic forum, registration, a web site and some controlled administrative access for my friends so they couldn't easily terrorize each other? I'm curious to know if I could start something like this on my own, and without much more than just my own server space, time, and creativity. I'm not looking to make a living out of this, its mostly just a way for me to more efficiently manage having several friends each wanting me to built or run a web site for them, and perhaps make some small residual income if a market exists. The Slashdot community represents such a broad swath of experience and expertise that I'd like to know how you would approach a project of this nature."

Comments

Plesk

[nhtshot]

1. Don't

But, if you insist..

Set up a simple box running Plesk. It automates most of the tasks of handling users, billing and maintenance. It also allows them to mange their own accounts.

Quick, simple

Re:Plesk

[paitre]

Run far, far away from Plesk.

It might simplify SOME things, but it sure as hell makes other things more difficult.

Re:Plesk

[12357bd]

Making a server run smoothly and securely on the big bad internet is an arduous task.

Not so hard, ADSL + NAT(on router) + Linux + Apache + mod-security + static content

Now, managing users... that's hard!

Re:Plesk

[SausageOfDoom]

Systems like that also do all the hard work for you - which is great if you know what you're doing, but otherwise when you run into a real problem, or if Plesk goes wrong, you have no idea what to do next.

I agree with the original comment of "dont", as far as setting up your own server at least.

Don't, because the market is full of bedroom hosts who don't know what they're doing.

Don't, because unless you're going into it seriously (and by that I mean investing time and money heavily, hiring enough staff to provide 24/7 support and decent SLAs, and charging appropriately serious money), the margins have to be so low to be competitive that you're losing money when the customer submits more than one ticket a year. Which they will do, because they've come to you, which means they don't know what they're doing.

But most of all, don't, because if you have to ask how to do it, you shouldn't be doing it. You really can't be going into this if you have so little understanding of the issues involved in running a server and the associated services that you need to ask. It's not fair on your paying customers, because when they have a problem, you won't be able to help.

If you want to resell space, do just that - go find a company dedicated to selling reseller accounts. They will give you a whitelabel reseller account and look after all the server issues themselves, leaving you free to pimp out the space.

If you do, just make sure you have an exit strategy, tied to some kind of dead mans switch (even if it's just leaving details with a friend) - I've heard of far too many resellers disappearing, leaving the customers unable to get access to their sites, and the resellers in a difficult position as they should have no direct contact with the end customers.

How to succeed in 10 easy steps

[name*censored*]

1) Line up a patsy
2) Get some matches/lighter/firestarters
3) Burn down all competing datacentres in your city
4) Set up a webserver company

Seriously though, it's an incredibly overcrowded market - if you have an idea on something new or innovative to offer, then by all means go for it. But as they say, there's nothing new under the sun, and you'd have much better luck trying to compete within a market that isn't so overcrowded. Professional encryption/sensitive data management perhaps?

Re:How to succeed in 10 easy steps

[stevey]

That's a great idea.

I originally setup xen-hosting selfishly because I wanted a decent root access level of hosting for myself, but didn't want to pay for a big machine.

Within a week I'd found enough users to bring the cost down to an acceptable level, primarily because a few people know me and trust me, but the intention was always there to document it fully and have people setting up similar things.

Two years on I'm not aware of anybody who's replicated the setup which is a real shame, I think there's a lot of space for a kind of "cooperative" hosting setup, each one with maybe 10 users.

Think *very* hard before you do this

[wrook]

My ex boss (in a *very* small company) did this for his friends/business associates. It's a royal PITA. Unless you *can* make a living off of it (and have a good business plan that convinces you it's feasible), I would recommend not doing it.

It sounds easy at first. How hard is it to just whack up a couple of simple web pages for a couple of buddies? Lots of us have one of our own and it takes almost no maintenance.

But what happens when your buddy starts to attract some undesirable attention? For example, maybe you buddy has a car dealership and just wants a quick and dirty website. But he pisses a script kiddie who then spends the next year trying to pull down your server.

Or what happens if the site goes down at 3 am and your buddy just *has* to have it up and running?

Or what happens if your buddy decides he just *has* to send emails from his website when someone clicks on something, and you discover that the package you are using has about a million vulnerabilities and you are now the biggest spam king in the US?

Honestly, it just sucks. Buddies who can't set up their own website are almost always unreasonable. And they will expect "professional service" even if you don't charge them. And they will bug you continuously for completely boneheaded things having to do with their site.

Unless you really don't like sleeping, I recommend backing away from this idea.

Re:Think *very* hard before you do this

[erikina]

Agree, although for me it has never got to the point of having my sleep interrupted. For a small amount of money ($70 - $200), I've setup a number of websites.

Their pages don't generate much traffic, so I said I'll host it for them and never discussed ongoing maintenance and changes. It's really been terrible. I get emailed (even phoned) all the time (esp. the guy that paid $70). Every week he'll wants something changed, or something modified.

I ended up drawing the line when he decided he wanted fancy roll-over menus instead of the current very functional one. I gave him a (large) quote just cause I was sick of it. We never spoke much after that.

With another, the business (who the site was for) was sold - so when I got in contact with the new owners. I told them, I'd continue to host it and charge a straight amount if he contacted me about the site and plus an hourly rate.

He thought it was very reasonable (after all, isn't it?) and actually has never bugged me once about anything.

The lesson is if you're going to setup a website, make sure you arrange the terms of ongoing maintenance. There's going to be a lot of it (esp. if you're doing it for free).

The other lesson would be don't do deals with friends. You'll both have completely different expectations of each other - and very well might ruin your friendship over it (unless you're a better person than me, and enjoy helping more than I do).

These days, if people want me to do any work - I tell them I don't know how. It really isn't worth it.

Here's what I do

[Anonymous Coward]

I run a gridserver account at Media Temple, $20/mo or $200/yr. Set up websites with 1-click apps like Drupal, Joomla, Wordpress or any other easy to use free PHP CMS. Wordpress is easy to modify and has a very large range of plugins and templates to work from. You can set up webmail access on MT servers as well as FTP and SSH and permissions for additional user access to your main server account if need be. Many other Hosting Companies have similar systems available or more. I have over 10 webpages on my server account and am barely scratching the resources so far.

If I make one webpage for a few hundred dollars, it pays my hosting for the year. Until I use 1/2 my resources, I have no need to upgrade so far.

Re:CPanel

[SimGuy]

Don't use cPanel. While it automates a lot, it also makes lots of arbitrary modifications to the operating system rendering it annoying for use for anything else. Also, it and Plesk install lots and lots of extra things you will never use, wasting disk space and RAM without major tweaking and opening plenty of potential points of intrusion.

I work at a web hosting company and I find InterWorx to be the best at doing a little automation without making a mess of everything.

That said, if you know how to use Linux, don't use a control panel. You'll find it easier to manage things yourself. Short of the MTA, these things are really rather easy to configure.

Control panels + advice

[QX-Mat]

Think. Stop.

It's more annoying than you might think. I've done it, all my friends have done it, my cousin's done it and our dog will be doing it soon.

Don't don't don't. It's a VASTLY under subscribed and overly competitive market. Once you think you're the best, and you're successful, you become too reliant on a core group of customers who won't last for ever.

There are reseller accounts available with lots of ISPs, but few are on a commission basis (ie: you're the one who has to cover your client's costs and invoice them). Flat fees are usually available to dedicated servers licensors @ £50/m+ - but the market is changing and I'm not at all surprised if they're cheaper.

Plesk - possibly the worst thing I've ever used. Convoluted backend I couldn't hack on to extend pop-before-smtp the way I wanted.
CPanel - the original but very costly 6 years ago when I last used it. Has some impressive addons
EnsimDirectAdmin - Not one I've used personally, but I hear its ok.
VHCS - Freeware. Never used it personally. But there are many OS projects and forks out there if you look hard enough ]

Cubepanel and BlueQuartz worth a mention.

Most of these project offer "lite" versions which are free for restricted personal use. The only major difference between the free and paid versions is that the latter has multi-user and reseller capabilities.

I'd recommending taking up a decent Linux or BSD distro with a proven track record of security fixes. "apt-get update" is sufficient for the home user, but realistically, you want to track purely security updates. Consider an enterprise OS (CentOS?!)

Matt

Here you go

[kylehase]

I'd start with 18 2x quad-core 2.3 GHz webservers each with 8 gigs of RAM, and 4 quad-core 2.3 GHz databases with 16 gigs of RAM.

Like everybody else says

[smoker2]

DON'T DO IT.
If you want to play with a server environment, get a mainstream linux distro and install it with all the web capabilities from the beginning. Then learn how to administrate it, install and modify php, learn apache control mechanisms, learn about chroot jails, even consider virtual servers (as in VMWare type of virtual).
Then if you're still up for it, rent a virtual solution from somebody else, and play with it a bit more. The costs to entry are very low, but there is almost no return.
Build your friends website by all means, but you're better off hosting it on a third parties hardware, and let them take the strain of running the hosting business. You still get admin access, but all the tools will already be in place.
I've been doing this for years (8+) and I get more hassle from the users than from the sites. You need a call centre just to explain to people that there is nothing wrong with the site, maybe their net connection is down. Or they're not getting their email, because their isp is blocking it, or the page doesn't look right because their browser is still caching it from last month etc etc etc.
I have someone who questioned why their site costs money each month to run. "To pay for the server" I replied. "Oh, I thought that once you had uploaded it, it was out there, on the internet" he replied. *Smacks hand to forehead*
Oh, and if they want to offer downloads, then make it clear that they will be charged for bandwidth, over and above any monthly fee. Do not give shell access out like candy, and don't allow anonymous ftp.
All in all, don't bother, unless you really are a masochist. By all means build sites for friends, but set them up with a host somewhere, and let them get on with it. 90% of people don't keep up with updates to sites anyway, and you get left with crap lying around on your server(s). I have 1 guy who bought a domain name through me around 6 years ago. He has never had a website for that domain. Every year I hit him for the renewal fees and he pays up, but it will never be a real functioning site. He is the best kind of customer. Beware of people who want stuff, especially those who think they know what they want.
Overall, realise that this is a huge subject with many many intricacies that you find as you go along. Do you really want to go down that road, or would it be much better to take the blue pill now and forget about it ?

Ran my own hosting company...Now i'm in catering!

[Anonymous Coward]

I did this for a couple of years. And the conclusions I came to were that:-

1. No money in simply hosting sites on a small scale
2. If you offer hosting, over sell your server capacity by at least a 1000% or you will never attract customers. They hardly ever use more than 1% of what they sign up for and won't sign up unless your storage/bandwidth offer is as ridiculous as everybody else.
3. Convert your hosting clients to higher value customers by offering web/graphic design opt-in email marketing services etc.
4. Either write your own services/management systems, keep your free software updated religiously or plan for the day when your free software is exploited and your server is owned by a couple of kids. When this happens your server will get null routed your customers will be angry and you will spend a couple of days at least recovering (damn OpenWebmail!)
5. Be prepared to answer support calls any time any place, no matter what crisis your life is in. Imagine troubleshooting a technical problem whilst on a stag weekend in Dublin!

If you can't do all this yourself, you need a pot of money to get a good team who can, but don't expect to make a good profit unless your added value is exceptional.

I got out of pure IT all together, I've found that it's far easier to get a traditional business off the ground and with the skills I've got my new company is light years ahead of the competition. How many small catering businesses do you know of that have 1TB File Server, there own dedicated web/mail server, asterisk PBX with VOIP/POTS lines etc, and a dedicated 24/7 tech support person with excellent dish washing skills?

Most of these responses are from losers and fools

[Anonymous Coward]

I am not going to tell you any details about how I've done it or with whom, etc.

I will say I am 75 years old, never took a course in any computer program, use Frontpage (oh the humanity! - but it does the job) most of the time, Photoshop, a good text editor (I like Alleycode) plus some Joomla, WordPress and a very good little shopping cart. I seek out and intently use various forums. I have one modest sized dedicated managed server based at a resource that gives great tech support, and another shared server account (likewise). I charge $240 per year prepaid (refund of unused portion anytime) and reasonable prices for site building.

I have clients in 7 countries. I'm not getting rich but it sure beats Social Security. I have a very satisfying retention rate; some clients are with me now for more than ten years.

How do I do this in the face of all the negatives, real and imagined above?

I give very serious personal service. My home phone is on my business cards.

I regularly study and monitor every client's web sites (some have as many as five sites) very carefully and am proactive in alerting them to issues that affect performance and value. I watch their traffic and the email flow and take action when I see a problem.

I deal only with the owner of the business or the top executive of the organization (I have a number of NGOs, some of them famous).

I avoid making presentations to potential clients whom I recognize early on as people who know the price of everything but the value of nothing. You can tell who they are because they only want to know the price, not discuss what their needs are or how my services will fulfill them. Let the discounters have them.

Based on years of experience, I never accept creative people of any kind as clients. That means, no writers, painters, performers, photographers, etc. - they are unteachable and surprisingly close-minded. Give me an inquiring, curious and engaqed business person any day.

When I screw up, I make sincere amends that build trust and loyalty. For example, when I failed to prevent an unintended domain expiration, I worked hard at recovery, got back the name for my client and gave him a free year of registration and hosting. He's been with me now for 6 years.

I never speak with anyone without giving them a business card. During a visit to any store or business or any casual encounter, I hand out a card. I give a free year of hosting to any existing client who sends me a new client.

In other words, bottom line, I work at getting customers and I work even harder at keeping them.

$15 per month buys you a shake-and-bake web host

[pvera]

I have had mine for about four years, and I started mostly because of the same reason. I needed a way to not spend too much of my personal time helping my friends deal with crappy web hosts and horrible domain registrars.

$15/month got me a reseller account with a very very nice host. When I was just their customer they treated me like royalty, I love their service. The $15 bought me space, bandwidth and access to WHM, the admin side of Cpanel. I was free to slice it up any way I wanted, and each of my customers would have their own cpanel.

The smart ones knew how to install whatever they wanted without the cpanel automation.

The rest were self reliant enough so cpanel was actually useful.

One of two idiots just couldn't pay attention if their lives depended on it, so it wouldn't matter which method they chose.

The second problem was domain names. Almost everyone was sick of the Netsol prices, but most got burned buying discounted domains elsewhere. Almost by accident I ran into a small domain wholesaler based in India, the reseller account was free so I decided to give them a try.

They are fantastic, four years later and hundreds of domain names purchased and NEVER a problem that could be blamed on them. Every single problem we have had could be traced back to a bad discount registrar trying to screw a customer out of transferring out.

The main problem with doing this is that customer support does take time. I made a point of always being available to anyone that used my services. I was always willing to IM or talk on the phone regardless of how stupid the issue was. Because of it my attrition rate is literally nonexistent.

By "accident" my previous boss overheard that I had my own domain registrar, and that my domains were less than $10 instead of the $35 he was paying.

Overnight I had nailed a 30+ domain customer.

The funny thing is that my hosting provider kept increasing the quotas for my plan, but kept the price frozen. That means I still pay my $15, but now I have 10x the disk space and at least 100x the bandwidth I started with 4 years ago. This allows me to be generous when a customer runs over bandwidth and I let it go "just this time."

They feel like I did them a favor, when all I did was use up a tiny bit of the extra capacity that was not drawing revenue.

Do I make money out of this? It pays itself, and I get some extra cash left. Many times friends ask me about cheap hosting for a family member, so I usually sell them the domain name and give them a small hosting account for free. If I botch something, I give them extra hosting, if it was a BIG screw up I give them a free domain (I don't remember giving away more than 2-3 domains in the past 4 years).

I also play with the allowances just to see how people react to it. For example, I may set a basic account so it has up to 5 parked domains (these point to the root of the account), 5 add-on domains (these are stand-alone within the account), etc. This costs me nothing but when they ask nice I say "just because you are a good customer ..." and bump it up to 10.

Then there are the miscreants. I had one customer overseas, I have no idea where he came from but he immediately bought the biggest plan I had, which was expensive and ate 25% of my resources (and I was not overselling, he DID get 25%). A week later his site auto-shut down due to bandwidth use.

He came back to complain, I explained to him what he had done. He asked how much. I honestly wanted him to go away, so I offered him to temporarily bump him by 1GB for the rest of the month for $100.

He said yes, and paid on the spot.

Less than a week later he was back. Same deal. Another $100.

$300 later he moved his site elsewhere and I never heard from him.

Another customer, also overseas, was good for about 5 domains per year and maybe $200/year worth of websites. The problem is that she IM'ed constantly, and for stupid crap. Every week she would reinstall what