Not sure if this sort of thing is allowable on here, so if not, just let me know and Iâll remove it.
In my role, I have been responsible for managing the various environments we use. One the systems I have been involved with was comprised of several applications, apis, databases etc. Each of these components are integrated in some way.
When one of these parts goes down, it can cause a knock-on effect on other components. When an issue is reported, it can take time to work out which component has gone down.
With that, one of tools I usually create and put in place is some sort of monitoring tool. In the past, Iâve written tailored solutions for the area Iâm working in.
I thought I ought to write something flexible, something that allows someone to quickly configure the components theyâd like to monitor. Iâm not claiming this is something unique or new. There are plenty of such open-source and commercial packages readily available. But, now and then, itâs nice to do something yourself and it gave me a chance to learn more about Node.js and React.
If youâre interested, I wrote a tool that will check websites, web-apps, APIs and servers are available and respond within the expected time. You can find it here. I should add, this is completely free and open source.
If you do try it, Iâd be grateful for any feedbackâŚ
Yeah, there are loads and I never claimed itâs an original idea.
I donât see why that shouldnât stop my writing my own though.
Iâve been at places where some of these commercial packages have been tried to be introduced. Quite often, theyâre intrusive in that JavaScript has to be embedded in webpages, or agents have to be installed on servers. In fact, it was the server agents that stopped adoption of one package as architecture governance wouldnât allow it.
I wanted to build something lightweight, unobtrusive and gave quick feedback. Plus, it gave me chance to learn more about Node and React.
The answer is âagentless application monitoringâ
Things like JMX, SNMP etc.
These âAgentsâ are already in every OS, Web Server, Database and in most popular applications. You should just enable them and use some tool to query them, thatâs it
Where are you going with this? I wrote something and you say I shouldnât have wrote it? Now going further with agentless solutions. Yeah, I know stuff is out there, I couldâve researched more and found something else instead of knocking up something myself. I answered a simple âwhat do you use q?â on here. Maybe Zabbix shouldnât have bothered because thereâs already solutions out there? Does it really matter that I wrote my own? I wanted a single dashboard that checked several asset types and I got it.
But, in a lot of places different groups manage different things. You canât get access to certain in-built monitoring tools because of access. I wrote something that met my needs, nothing more.
Oh, it looks like you have a site aggregating remote job opportunities.
Thatâs original. Why bother? Why go to effort putting something together when thereâs loads of job sites, remote and site based?
Why? Because you wanted to? You feel like like youâd get something out of it? Someone else may benefit from it?
In my original post I acknowledged, âThere are plenty of such open-source and commercial packages readily available. But, now and then, itâs nice to do something yourself and it gave me a chance to learn more about Node.js and React.â, yet you chose to list existing solutions as though I wasnât aware that my idea wasnât original and suggest âwhy botherâ.
Only if you can call balanced nodes directly. For example, if you have an EU node, a US node and global node which does nothing more that route to EU and US. Then, you need to monitor all three. Just monitoring the global node wonât tell you which underlying node had gone down.