Prohashing Mining Pool Forums

Hello,

I was looking at ProHashing's Alerts, and would like to make a recommendation to add an alert for a miner that has been down for X minutes. I have been trying to use the alerts to monitor my rigs, but found each of them don't quite hit the mark, when all I really want to know is if one of them stops working.

In my personal situation I have two larger rigs (multiple newer cards), one small miner (a single 2060), and my little Chia plot. I also on occasion test miner software on my own system.

Profitability, Total hashrate, and accepted shares won't help. These all vary over time, so you need to pad the range you are trying to monitor. Using the range, if that little 2060 dropped it would still stay within the range most of the time.

Named worker count is closer to the mark, but if your trying to get larger institutions to join ProHashing, having to come in and adjust that all the may be a little annoying. In my particular case I had a minimum of 3, which is fine most the time, but I happened to be running my test rig when another went down, so was still at 3 and didn't realize one of my normal ones was down for a while.

So what I would like to see is an option for setting an alert when a worker goes down for some amount of time. Allowing us to pick a range of time would allow for reboots and restarts. If the worker does not come back after that time limit, then we get a single alert (not a constant barrage, incase we have chosen to shut it down).

From a technical standpoint this may be a challenge since it looks like pro-hashing assigns each worker a random ID on the back end, so would be hard to know if it had restarted. In this case I think it would be fine to say the alerts only work with named workers, that way you can watch for the same worker name coming back online. Though if people gave two workers the same name, then it would also cause a problem, but that would be on us (the users) to resolve.

As I said, I like all the other alert options and can see uses for them, but really the most basic need it to make sure every one of our workers is up and running, and the existing options don't always do that for us.

Thinking about this over breakfast, was thinking of three :

Named worker offline - This would be, as mentioned in the original post, for named workers going offline for a configurable amount of time.

Named group offline - This would allow us to know if an entire group was offline, like "Grandma's house" for a configurable amount of time.

Worker offline - This would trigger any time any worker (even unnamed) went offline, without a delay.

I added the last suggestion which would not watch for the worker coming back online, but any time any worker went down. People may want this, so they are aware even when a miner restarts (allowing us to see if one becomes unstable). This would then also work with unnamed workers, since you wouldn't have to track if it came back online, and just sends out a warning that "Hey, something just disconnected"

The difference between the first (Named worker offline) and the last (Worker offline) would be that you could reduce alerts by using the named one, and if your rig can recover itself, then you do not need to get alerts so long as it comes back within the time specified. The latter would trigger every time, for people too lazy to set names (but then have to try to figure out which one of their rigs died) or wanted alerts even at restarts.

I created a ticket for Steven to look into this.