At Jitbit only two people have access to the entire company infrastructure - me (Max) and Alex. If something goes terribly wrong and the team can’t handle it, one of us has to get online and fix it. And we learned how to handle it pretty well over the years. Like everyone else, we do have occasional downtimes, but we never had a major outage in 10 years. The longest one we had lasted two hours and that one was planned maintenance taking longer than expected.
Believe it or not, we also take vacations occasionally. Here is how we do it:
While one of us is away on a trip the other one stays home and online (have I mentioned we are a remote-only company and do not have an office?). We just have to plan our vacations ahead. We usually are pretty good at it and our schedule has been more or less stable:
Alex takes a couple of family vacations, a mounting biking trip and a couple of conferences every year. I usually go snowboarding once or twice per season, take multiple weekend-long trips to various places in Europe and one proper vacation.
Sometimes we are both away for short periods though. Like I'm writing this on a plane while Alex is offline camping in the woods with the family. So if something breaks right this moment no one will be able to fix it for a couple of hours. Here's hoping that airplane wi-fi becomes more common and stable.
Over the years when something important broke down we added a monitoring system that alerts us about any issues with the component that died. By now every vital part of our SaaS Helpdesk is covered by some kind of a watchdog and all of those systems have backups as well.
When something goes wrong both of us get four alerts - an email, a text message, a Telegram bot message, and a Slack alert. Other employees also see these alerts in Slack and can call us if none of us reacts quickly.
In addition to the critical alerts for specific modules, we have general ones. Like, "the CPU load on this server has been really high for the last 10 minutes" (powered by AWS Cloudwatch). In that case, we go to that server to see what's going on.
We also have automatic-recovery features built-in. For example, if we know that something can be quickly fixed by recycling the web-app service we just do it in code automatically.
All those systems are there for our peace of mind. Instead of checking your phone every 30 minutes (I still do it anyway) you know exactly when something goes wrong.
You want to maximize the time you are online. There are a lot of travel SIM cards out there and most of them are pretty cheap. You can even buy it on Amazon for the country you’re travelling to. You may also check with your cellular provider and see if they have some discounted roaming options. They probably do. Ten years ago cellular data abroad was ridiculously expensive. Now there is no reason not to be connected the entire time you are traveling the world.
Do not ever rely on free wi-fi. I've stayed at hundreds of different hotels, Airbnb's, etc. all over the world. I had a good, stable wi-fi connection probably about 5 times. Your phone connection is almost always going to be better.
Currently, I'm returning from a trip to Göteborg. I was visiting a friend and he took me disk golfing (I didn't know that was a thing either). I spent three hours of non-stop hiking, going up and down hills and throwing frisbees into baskets. Lots of fun, but I had my 8-kilogram backpack with me the entire time.
I took my backpack to hundreds of museums, castles, fortresses, lakes, mountain roads, and long biking trips. I will probably have my laptop with me if I decide to climb Mount Everest. When I leave the house to get a haircut I still have the laptop with me.
I also carry a high capacity power bank with a USB-C port that can charge my laptop if I ran out of charge.
The things listed above may sound completely paranoid to you. And they are to an extent. However, we believe that keeping our company small gives us a big edge - we are very flexible and fast and we would like to keep it that way. Even if it comes with some extra overhead.