Off-site status information for cornerhost.com.

2/22/2003

control panel fixed

Okay, I've gotten the control panel working again (as far as I can tell)...The innercircle archives should be fixed later on today.

2/20/2003

ugh.. what a mess

Well, I moved three accounts tonight. Two went [fairly] smoothly, and the third one was a disaster.

Luckily, the third one was mine. Unluckily, it broke ALL KINDS of stuff, like the control panel, innercircle list, and I'm sure plenty of other things I haven't noticed yet. I think all my email addresses were down for about two hours. Anyway, I'm probably going to be behind on things all day tomorrow while I clean this up. (And no, unfortunately, I can't put things back the way they were... :/ )

2/19/2003

moving stuff around

I'm moving the control panel tonight... It may be down for a while, or give some nasty errors.

2/18/2003

staying put

Rackspace and I had a nice little chat, and the the upshot is that mercury, scandium, and titanium are all staying put.

However, I will move 10 or so of the biggest and most active accounts to chromium starting tomorrow, and will probably move the MySQL onto a dedicated machine this weekend.

I may wind up rearraning the file structure on mercury and upgrading some of the software, but I'll save that project for another day. :)

2/17/2003

mx records

I have added MX records for all domains on my nameservers. These are records that say where to send your mail. Since they just point to the same server your domain is on, nothing whatsoever should change. [Except that it fixes a problem with a tiny handful of outside mail servers.]

But if by chance your mail is suddenly broken, please let me know!

2/16/2003

the big change

I'm wearing my CEO hat today.

I'm about to make a big change with Cornerhost. It's something I've been contemplating for several months now, and I think it's time to finally take action.

This email will explain exactly what I have in mind, and what's going to happen. My apologies for the length, but I wanted to be clear about *why* I'm making this decision, and not just announce it without explaining the underlying motivations.

So here goes.


Lately, Cornerhost hasn't been as good as I'd like it to be. There are so many things I want to make better here, but it seems like I spend 60% just handling the day to day work of running the business, and 30% of the time putting out fires.

See, there are plenty of things I would do differently if I'd started cornerhost today - things that would make for a knock-your-socks-off FANTASTIC web hosting company, instead of just a pretty good hosting company with friendly support but recurring server problems.

But to do these things, I need to be an entrepreneur and a programmer, not a sysadmin and customer service rep. Since I can't be both, I'm making it my #1 business goal to hire someone to replace me! :)

The trouble is I can't afford to hire help yet. Cornerhost is profitable and gives me a small paycheck every month, but I racked up a good five figures of personal debt to get here. I have no intention of borrowing another dime, and I'm not ready to approach investors.

So the way I see it, I have three options:

  • focus on getting by, save my pennies, and continue to slowly grow a mediocre business, until I have the resources to really take off. [This seems to be what I'm doing now. I don't like it.]
  • focus on sales and marketing, knowing that increased sales will just add to my workload, but if I keep pushing I'll be able to hire someone sooner. [This will just stress me out, and it isn't fair to you guys.]
  • focus on the few area of my business that can MOST improve profits, while making the business easier to run.

I think option three is the most sensible approach long term. But I also think it's going to be a pain in the butt for a lot of people, which is why I've tried to explain my thinking here.

Basically, option three means rebuilding mercury and scandium. Starting from scratch, incorporating the lessons I've learned over the past year and a half.

Here is what I have in mind: move away from Rackspace.

I no longer rely on Rackspace's "fanatical support". I haven't opened a support ticket in probably nine months, except for reboots. Even then, their service mostly gets in my way. I love their reliability, but I'm paying an arm and a leg for support I don't need.

Luckily, Rackspace has a "kid sister" -- ServerBeach. Serverbeach was founded by one of Rackspace's founders, and Rackspace invests heavily in it. I've been using them for several months now [helium and vanadium are already hosted there], and am very happy with them.

Moving mercury and scandium (and eventually titanium) to ServerBeach should save my business about $1000 a month. Between that money and my current cashflow, I should have no problem hiring another "me" to answer email.

Just as importantly, moving the servers gives me an opportunity to solve the overcrowding problem on these two servers. With SeverBeach's pricing plans, I can replace mercury and scandium with five machines and still be able to hire help. Those machines would be:

  • three web hosting servers
  • one dedicated database server
  • one cvs server [which I set up this past weekend]

So the upside of all this is that when it's all over, your accounts will run better [fewer accounts per machine, and no running out of diskspace on those 60GB drives], you'll have a dedicated support person, and I'll be able to focus on the good stuff - like a WAY better control panel, an affiliate/reseller program, a knowledge base, a backup system, autopayment for credit cards, suEXEC for mercury users, and *MAYBE* even iCal for you Mac folks. :)

But to get there, we've got to pay a price. The price is a whole lot of work on my part, and a certain amount of inconvenience for every single user on those two machines. Mercury may be especially troublesome, because the directory tree also needs to be changed (for reasons related to suEXEC), and this may require changes to some people's scripts.

Of course, I'm going to attend to each site individually. The machines are under contract until April, and I can go month to month afterwards if need be. I'm going to take the time to do this right and make sure it goes as smoothly as possible for everyone.

I don't yet have a detailed schedule, so I can't say who will be moving when. But, roughly, the process should take two or three months to complete, and it'll go something like this:

  • move hydrogen (secure.sabren.com) over, because it's easy and I can do it without affecting anyone.
  • move db.sabren.com to a new, dedicated machine
  • move the highest bandwidth sites from mercury and scandium over to new machines
  • move everyone else over from mercury
  • move everyone else over from scandium
  • use monthly savings to hire first employee
  • possibly move titanium over in september

So there you go.

Once I've got the plan nailed down better, I'll send out an email to everyone, and then start contacting people individually to let you know when your site will be moved and discuss any special needs you may have.

Meanwhile, I'm posting this on innercircle first so I can get some feedback here. The decision is made, but I'm sure there's plenty I can do to make the change as easy as possible, and I'm open to suggestions.

Blog Archive