Because I run my node on a computer I don’t use a lot, I don’t even have the computer connected to a monitor.
Would there be a way to create a secure and simple interface where we enter our node’s public key, and the interface just tells us whether the node is online or not?
That way, we could just check our node’s connectivity from any other computer. If we find out it’s not online, we can then specifically try to troubleshoot it and reconnect it.
I know there are existing instructions for how to access the Storj dashboard from any other computer, but that seems overkill for my purposes.
All I need is a quick and easy, “Yep, this node is online” or “No, this node is not online.”
The satellites will send you email about suspension when your online score is dropped too much, so it’s already too late. It’s better to use an external monitoring for that.
Thanks @Alexey! I’m a newbie to being a node operator (I set up my node about a week or so ago).
Several days ago my node stopped working for whatever reason, and I didn’t know about it because I hadn’t set up UptimeRobot yet.
Long story short: the node was down for several days. I discovered it when I finally plugged in a monitor to check on it. Once I started the node again, I realized my percent uptime was down to like 20% or something.
Are there plans to maybe introduce some kind of trial period for new node operators, where we aren’t paid yet but our percent uptime is also not negatively affected?
That way, we can use this trial period to work out all the kinks in how to operate a reliable node (e.g. install UptimeRobot, make sure OS is not logging user out after period of inactivity, etc).
Or is a low initial uptime score not a big deal and easily recoverable over time?
Online scores drop very quickly when the node is very young.
In theory though, AFAIK, you have approximately 10 days of grace period if your node gets offline before getting suspended.
Your online score will recover within 30 days if you’re now steadily online from now on.
None that I’m aware of, but you can try things out on a real node anyway, because nodes make so little revenues at first that losing a node during its first weeks really isn’t a big deal. You just have to start a new one
The real frustrating issue is losing a node that got the time to gain data and reputation (say a 12 months old node for instance). So yeah, the best thing to do is to keep an eye on it so it stays online (uptimerobot is perfect for that), and have good scores.
The most important one is the audit score: if it starts dropping, red alert as things can go out of control very quickly and could lead to disqualification.
I believe it’s even better than that. Last I heard nodes don’t get suspended or disqualified in the first 30 days. When you do get suspended, you get 7 days to fix your issue, followed by a 30 days monitoring period of which you are allowed to be offline for 12 days. While your score is below 60% you won’t get any incoming traffic though, but your node will survive.
I think there is plenty of slack in there that anyone who is trying to make it work in those first months would easily survive that. Just make sure you’re not offline for 30 days straight as that would lead to instant disqualification.
as things can go out of control very quickly and could lead to disqualification.