It clearly could have been much worse. In an area where some customers lost power, in the middle of a very cold winter, for as long as a week, I was fortunate that my power problems lasted only 3 days. One day of complete outage followed by 36 hours of intermittent, but primarily available power, is better than being one of the customers who just got their power back after more than a week.
On the plus side, it was nice to see that my, previously untested by real-world conditions, UPS systems performed as advertised. Graceful shutdowns and immunity to the blipping power issues kept everything operational and allowed me to make the choice to leave systems down until the power was fully restored without any damaged equipment or lost data. And throughout the power failures, my Internet provider stayed up and running, justifying the minimal expense of the UPS on the external router.
Initially, I had put the router on a UPS because I switched to VoIP phone services almost a decade ago, and with no traditional landline investing in a UPS that could keep the phones up and running made sense. But as soon as the power was reliably restored I was faced with a new problem; my ISP connectivity went down. Not only did it go down, but it stayed down for the next four days.
To some, this might seem a minor inconvenience, but my business is almost entirely virtual. I connect to clients, write, research, and work sitting at a computer, using the Internet as my resource connection and primary tool for getting work done. So being without a connection meant that a lot of work couldn’t get done.
My provider (Comcast) was of little help; each day they reported that the outage would be corrected by that evening, and each day the updated status reflected it would be fixed the next day. This is actually the first extended outage I had faced in more than a decade of service with this provider, and when compared to people facing freezing temperatures and miserable weather without power, my lack of Internet service isn’t that big a deal. But it did put be behind in getting work done for myself and my clients.
The nature of the outage made it impractical for me to stop by the local coffee shop and work on my notebook. They had opened their doors to people without power and were pretty busy with folks stopping by to recharge mobile devices and keep warm. And I’m always a bit leery about accessing client sites using a public Internet connection, regardless of the security on my end. So this meant using one of my 4G devices as a wireless hotspot to connect my home office network to the outside world.
On the surface, this seemed like a simple solution. I have a decent 4G service and reasonable bandwidth availability. I use a wireless bridge to connect my wired office network to the net, so reconfiguring it was not a major issue. But after the first day I realized that it would be necessary to make some changes to the network. Working strictly from a mobile device such as a tablet simply doesn’t fit the way I work.
Since I make use of cloud services, there is a fair amount of network traffic in the background and I found that I was burning through about 500 MB a day with my normal setup even though I was trying not to do things I knew would suck up the bandwidth of my 4G data package (For example, I normally stream a news channel while I’m working). While a large percentage of that data traffic was email, a significant amount was the replication service I use between my NAS and cloud backup storage, as well as inbound traffic for clients accessing resources on that NAS system.
Normally, I don’t pay attention to the amount of network traffic I generate. It’s not capped by my provider (though at one point they were limiting users to 250 GB per month) and it doesn’t impact my work but now I found myself with a potentially very large mobile phone bill should the Internet outage extend for a significant period. What this meant was that I needed to figure out what was causing the traffic and determine if it was absolutely essential to my business.
Because my high-value data backup system is disk-disk-cloud I was relatively comfortable disabling the final step for what I presumed would be a short period of time. That alone significantly reduced my network traffic. I also made sure I was working strictly with local copies of data and not making demands on the connection for cloud-hosted resources.
The outage did make me rethink the way I work; without really meaning to. I had been integrating cloud-based services into my daily business process. These outages made it clear just how critical my Internet connection had become to my line of business operation. In my case, without an Internet connection I was basically shut down. So to prevent this in the future, I’ve been working on restructuring some aspects of my business workflow so that temporary interruptions don’t stop me from getting my work done.
With a predicted blizzard hitting tonight and possibly again in a few days, these plans may get tested sooner than I had expected. And it prompts me to pose the question to you; How much of your business workflow depends on a reliable Internet connection?