how-to-deploy-ready-to-use-ml-packs-for-

How to Deploy Ready-To-Use ML Packs for IT Monitoring in 2 Easy Steps

February 16, 2021

Organizations everywhere are discovering the joys to have ready to use ML Packs for many technologies: Collect, Analytic, Trending, Dashboards, Alerting. And they can be enabled with just a few clicks! 

In recent years, it’s become simpler to use opensource and built yourself whatever you need to monitor whatever you want. Nonetheless, there is one drawback to this DIY approach, and that is time. It’s a lot of time to develop this, and even more time to maintain this. Since the beginning, our mission is to help you achieve a maximum of visibility and enable best-practice monitoring thru advanced analytic and alerting.  

We have received praise for this from our customers. Recently, one of them told us that quick deployment of Unryo Monitor was the biggest advantage for his company. For this customer that is using the Unryo Cloud edition (SaaS), it was less than 5 minutes to have his datacenter collector in place. And then, from the portal, he enabled full monitoring of his VMware infrastructure in less than 1 minute!

Let’s demonstrate how you can do this as well. By the end of this short demo, you will also become a fan of Unryo Monitor – guaranteed!

New Customer, Installation Step 1/2 

From the Unryo Console at portal.unryo.com , you click +Deploy to deploy your collector.

2021-01-29 14_13_P1.png

Next, select a name for the new collector. You can install as many collectors as you want.

2021-01-29 12_23_44-Unryō.png

Indeed, the most popular deployment type for our collector now is a docker-compose file. That make that collector easy to install, easy to auto-upgrade. This file is all preconfigured, most of our customers don’t change anything in that file. But you have the flexibility to adjust the docker parameters if you want to.  

 

New Customer, Installation Step 2/2

Run docker-compose on your server to install that pre-configured collector. Instantly that collector becomes visible in the management portal : 

2021-01-29 13_37_13-Unryō.png

Believe it or not – you have nothing else to install! Everything else you want to do with Unryo Monitor will be achieved playing with the Unryo Monitor web portal ! So, close the management portal (portal.unryo.com) and from now on, just focus on YourCompany.my.unryo.io

This is your day-to-day Unryo Monitoring Portal. Since you are admin, you can access the “Configuration Management” section of the UI and configure anything you need. 

2021-01-29 13_45_25-Unryō Portal.png
 

For that blog, suppose your objective is to monitor your vSphere environment. Here the steps: 

2021-01-29 13_48_28-Unryō Portal.png

Search within the hundreds of ready-to-use MLPack the name of the technology you want to monitor. In our case, let’s search for “vSphere”: 

2021-01-29 13_49_17-Unryō Portal.png

Then follow the instruction. Provide your vCenter URL and Read-Only Credentials. 

2021-01-29 13_51_25-Unryō Portal.png
 

Mission Complete!

From here, you have all the dashboards, best-practice alerts, trending, analytics engine starting to do their jobs! If you want you can visit the “Alert Definition” section to adjust alerting to your need, but seriously, default parameters are probably exactly what you need.  

Maybe you want to configure external “Alert Channels”? By default, all alerts are always visible in our built-in alert management console. You can acknowledge or take ownership of these alerts. But if you want them sent outside, feel free to enable any of the 16 channels! 

2021-01-29 13_56_11-Unryō Portal.png

Let's see the results! Let's try search and filter on the main page:

  • I can see all my vSphere components and can drill-down for full details on any of them. 

  • I can see that already some of the best-practice analytic rules have created some low severity warning. 

2021-01-29 14_13_P2.png

There are LOT of reports ready to consume, but this blog would be too long to show them all. Details about these reports are out of the scope of this blog also for now, let’s just click on that hypervisor where some memory contention occurred a while ago: 

We are not showing the full report page here. Nonetheless, in the below screen, you can see a cpu, memory and network peak happening around midnight…. And you can see that it's all coming from the VM called “Bronze NAS Service”.   

2021-01-29 14_13_22-Unryō Portal.png

There are so many technologies you can start monitoring like this, in a few clicks. Syslog management, application tracing, IOT sensors, and much more … For this and many more reasons, Unryo Monitor is a next-gen performance monitoring tool for next-gen organizations!

 

 What Next for Unryo and you?