What is the best way to get Server Stats for my custom status pages?

DianTama

Hi!

So recently i just used UptimeRobot monitoring for my website & server. but it doesn't have server stats like hetrixtools does.

And I'm just thinking, what if I use UptimeRobot API and then just get the server stats from my server directly.

But the problem is I don't know which tools to use to get the server stats.
Do you guys have any info about this? or the best way to implement this?

I have designed the StatusPage (it's highly inspired by the UptimeRobot status page ) I just added something missing from them. You can take a look here : Status Page UI Design

Cheers,
Thank You!

definitelyliam

You could do it similarly to how HetrixTools and a couple of others do it: create an agent.

Depending on your status page (if you're going to use the UptimeRobot API you're probably a developer writing your own solution?), you could write an agent that "calls home" at an interval that contains your server's stats at that specific point in time.

Your receiving API ("home") then receives these results and aggregates them so you can make graphs and logs.

A bash script + a cron job that calls it every X minutes is probably a good start for something simple.

DianTama

@definitelyliam said:
You could do it similarly to how HetrixTools and a couple of others do it: create an agent.

Depending on your status page (if you're going to use the UptimeRobot API you're probably a developer writing your own solution?), you could write an agent that "calls home" at an interval that contains your server's stats at that specific point in time.

Your receiving API ("home") then receives these results and aggregates them so you can make graphs and logs.

A bash script + a cron job that calls it every X minutes is probably a good start for something simple.

Ah, that's one thing, after seeing hetrixtools agent bash script I got a concept on how to achieve this.

But the thing is if 1 server post 1 data every minute there could be 1440 data for 1 day. Doesn't it too much data being posted for a year maybe?

Edit : nvm, I got it now

definitelyliam

@DianTama said:

@definitelyliam said:
You could do it similarly to how HetrixTools and a couple of others do it: create an agent.

Depending on your status page (if you're going to use the UptimeRobot API you're probably a developer writing your own solution?), you could write an agent that "calls home" at an interval that contains your server's stats at that specific point in time.

Your receiving API ("home") then receives these results and aggregates them so you can make graphs and logs.

A bash script + a cron job that calls it every X minutes is probably a good start for something simple.

Ah, that's one thing, after seeing hetrixtools agent bash script I got a concept on how to achieve this.

But the thing is if 1 server post 1 data every minute there could be 1440 data for 1 day. Doesn't it too much data being posted for a year maybe?

Edit : nvm, I got it now

Well, you could override the current values so you only get one entry.

For me, I'd rather keep the entire 1440 data for 1 day. In today's cheap compute solutions 1440 rows a day isn't a big deal. It's also a huge factor when computing uptime statistics anyway.