Inspiring!
How do you handle on-call / customer support, particularly around vacations?
(In other words, if you want to go away for awhile, how do you make sure any outages get resolved?)
How do you handle on-call / customer support, particularly around vacations?
(In other words, if you want to go away for awhile, how do you make sure any outages get resolved?)
Another point is if you’re also your own product manager and you designed every feature yourself, I think there’s a tendency to think about all the eventualities for each feature more thoroughly, so your code and product is kind of complete - thus you actually get fewer bugs. In my experience most issues in software come because the engineer misunderstood the requirements that someone else wrote anyway.
How do you do to get alerted of your service going down when:
- you are in the wilderness, ouside or with little cell coverage
- drunk and dancing in a wedding with music at full blast
These are just 2 small examples out of many others. Also it means you need to stay connected when taking a plane, be able to stop and/or swap drivers if your are in a long driving trip so that you can fix your service in a rest area or while your partner is driving, etc.
I am pretty sure outages are very rare but if that happens the day your are out of cell coverage and unable to react, you might lose a lot of trust from your customers.
I am also suprised he relies on only one hosting service. I would have thought you might not want to have everything in the same basket.
If I am drunk dancing in the woods as a solo operator no one is doing config changes on my servers.
The remaining 5% downtime like internet connection to the server facility, solar flares, UFO taking over the world - I would not able to do anything about anyway, would have to wait until it goes away.
The most common issues I have with long term projects are small code/config changes (#1 by a ridiculous degree), then way below is drives filling up and then way below that is actual hardware failure.
Drives filling up are often the most annoying to fix as they cause very odd user facing problems. Just fixed this week an old wordpress site that I spent over 2 hours trying to fix, trying everything, before I realised the mySQL database was full. It turns out if a mysql db is full the provider revokes insert privs on the user and wp login will just go in a loop back to the login page if it can't write to the database. No errors on the UI :(.
Cloudflare is again down and your service is not available for an hour? You cant do anything. And nobody unsubscribes.
Most downtimes are because another service is down and you cant do anything.
Sometimes its really your issue. Your service is down for an entire day? Some customers may be pissed at the moment. But in the end they dont care if you are not a critical service - like this one.
Learning: Downtimes are not an issue. When such a message pops up just ignore it and look at it when you have time. This is strange first as we try to be perfect. But the world will continue to turn.
This is so true. To me, finishing writing the code means releasable code. I have done the testing along the way.
It runs on namecheap shared hosting.
If it goes down, I trust namecheap to fix it asap. If it goes down and someone called me, I don't really know what I would do anyway....
In 6 years it has never been a problem.
PHP and jQuery. I use phpmyadmin via cpanel to manage the database.
- No phone number for support
- Extensive FAQs to let people help themselves first
- Vacation: take laptop along, check emails every couple days (but I haven’t tried remote vacation without internet ever)
Generally speaking: outsiders vastly overestimate the support burden of a one-man business. Maybe I’m lucky, but I only receive 2-3 emails per day with 25k active users.
In reality if you don't deploy new version or don't change config or don't post your product to HN server will be there simply running.
In big companies you might have bunch of people doing config changes all the time on different levels and you might never know when someone will break something you rely on.
They probably set up a HealthChecks.io alert.
I'd love to travel to a remote island, or do a 2 week hike out of cell service but it's difficult. The odds are incredibly low that downtime occurs in that window, but Murphy's Law and my anxiety won't allow it. The pros greatly outweigh the cons though. While I can't do those remote trips, I can still travel wherever else and just ignore things unless there's a downtime alert or an urgent support ticket.
Then just remote into systems, and fix issues. He’s pretty much “on call” 24/7/365
Also there have been times when monitoring alerts start blaring at 3AM, and there's no more sleep that night. Thankfully does not happen very often :-)