Migrating VPS Providers and Why

A while back, I talked about legacy hosting providers and the value they hold (or in many cases, don’t). I touted the benefits of procuring your own Virtual Private Server (VPS) or if you only have one or two sites, finding someone like me that can host your stuff reliably with all the bells and whistles for cheap. These old legacy guys bank on you being too lazy to move your stuff so they can continue to charge you like they did in 1997.

I still contend that VPS is the way to go, but as mentioned above, reliability is key. Your site is no good to you if it is down all the time and if you run an eCommerce site, no amount of sales prices, SEO, and great product can contend with not being up. This is where we started running into problems with our old host and I’d like to share the last year with you to explain why I moved and why you should too if you’re having the same issues with your VPS provider.

I would also like to note that I will be critical of the old provider and name names, and I am not receiving anything from the new provider to write this. I am writing to share my experience as a customer of two competing services and show you that it is not hard to migrate VPS services, even for a non-server-admin like myself.

About three years ago, I tagged onto a friend’s VPS and DNS services for a split MX record on myTooq.com. As things progressed and I started building sites, I essentially took over his VPS with him running a few projects there still. This VPS was with A Small Orange based out of Atlanta, Georgia. They were excellent as far reliability and customer support went. I mean, I was (am) an idiot when it comes to a lot of this stuff and I could send an email and get their help on anything. They wold respond, usually within minutes and things would doing what I needed them to be doing in no time. Roughly two and half years ago, my buddy was looking at their site pricing and found that he was being charged about 25% more than the current rates for a more powerful VPS. When he asked them about it, he was basically told that it was his duty to ensure that he kept his eye open for better pricing (which I agree with).

We discussed things and migrated to their “Cloud VPS” service and things were pretty great for a while. We had a faster rig, lower price, and ASO migrated everything for us with no major issues. Then, in January of this year, I received an email from a client that his site was down along with his email. This was on a Friday afternoon, and a quick check showed that everybody was down, which I think at the time was a total of 32 sites, of which about 26/27 were externally facing client sites. I fired off an email to support as I had done several times in the past and didn’t hear back for a little longer than normal. To make a long story short, ASO had farmed out this part of their business to another company (I think HostGator), and they had a massive failure. I woke up Saturday morning to an email stating that all backups had been lost along with the data on the server. We relied on their backups for our data (I know now this is very bad practice). I did an inventory of the locally stored assets I had (I had a few full site backups and a lot of content backed up) and started contacting customers to share the bad news. I was essentially going to have to rebuild thirty sites as a fast I could and I was devastated at the prospect. By Saturday night, I had a few site back online and was stressed out and exhausted. Sunday resulted in a call with the possibility that they would be able to recover at least some of the data and luckily by Sunday evening, almost all of our sites were back up and running (the rest restored overnight). Not all of their clients were so lucky.

As it turns out, ASO had built their own system and had planed to start migrating customers on the Monday and had the failure happen a week later, it likely would not have affected more than a handful of people. I talked with a few of the ASO execs and decided to give them another shot (and commended the grace under fire of their support staff). I also provided some feedback on communicating more openly about what was happening. Obviously, they don’t want to announce that kind of failure to the world, but they were not thorough with their email lists and important communications were being missed. Our sites had been deployed to the new in-house cluster on the Sunday night and we were set to go.

Fast forward a few days and things are down again. Turns out there was a hardware failure of a non-destructive kind (I could go check my emails to see what it was, but the specifics aren’t huge here).  They get things fixed and things are up and running. Then a couple of weeks later, something else goes. Then again, and again. Downtime is not usually massive, but anywhere from 5 minutes to half-an-hour at a time every 10-15 days is still unacceptable (and this is when I or a client would notice). I installed Copperegg for monitoring and found that things were down far too often to be considered reliable anymore and it was time to change. I set up an account at BigBrainGlobal.com after reading reviews on a web hosting forum and seeing many reviews. They migrated everything for me, but I wasn’t ready to make the full switch. I developed new sites there and left the old sites with ASO for several months.  The BBG VPS also had better specs (2x cores, 4X RAM, 3.5 X storage, & more monthly bandwidth than I can dream of using) and an easy to find lifetime 25% off deal.

Since the summer, I have not seen an unplanned outage on BBG. There was a maintenance outage with a few minutes of downtime, but it was done late at night and well communicated and literally 5 minutes long. They sent an email warning about Hurricane Sandy and that they had prepared as best they could, but their data centre was in the storm zone and may be affected. Luckily, their preparations prevented any issues and all was fine (staff was safe too). During the storm, they sent out one or two updates. When the danger had passed they sent out an email. I knew what to expect while they were in the middle of a hurricane.

Last week, I remigrated all of the sites from ASO to BBG in an afternoon and with no major issues (had to remigrate to catch changes made in the 5 months of BBG running next to ASO). I had a few things to fix up with some certificates and site references, but essentially it was pain free. Things are now running on a much beefier server for a lower cost and I don’t have to field calls every ten days about things being down. I will say that BBG’s support does not respond quite as quickly as ASO’s does, but it is certainly still something I’d consider reasonable (fast by most company’s standards). They also seem to know how to solve things quickly and permanently.

I think that A Small Orange is undergoing some growing pains and suffering their own success. Their infrastructure is not matured and reliable, communication procedures are going more corporate (cautious and worried about legal over accessible, timely, and direct), and lots of new support staff aren’t yet as knowledgeable as their senior peers. I will still say to this day that ASO support, even newer staff, are some of the best support agents when it comes to speed, empathy, and taking ownership of a problem. I wish them, and ASO, the best as they figure out this growth spurt. As to Big Brain Global, I look forward to reliable service and growing my business with your services.

Migration is pretty easy. If your host isn’t cutting it, make the call and make a change. A small amount of pain to migrate offers a chance to receive better service and uptime, but you can also find better pricing and juicier servers in the process.

Advertisements