<![CDATA[Mango hardware configuration]]>Hello everbody.
I am new on this post, I have a question about Mango hardware configuration.
In a new project I should configure Mango to check ten remote sites, each of which has one hundred variables to be measured. What hardware features must have the mango central server?
Best regards, Luigi

]]>https://forum.mango-os.com/topic/2351/mango-hardware-configurationRSS for NodeWed, 15 Jul 2026 04:37:28 GMTMon, 18 Jul 2016 18:11:16 GMT60<![CDATA[Reply to Mango hardware configuration on Mon, 25 Jul 2016 14:06:29 GMT]]>We're running about 200 points on a Virtual machine with 1 CPU and 4GB of RAM.

Normal system usage is about 1.5-2% CPU and 1.5GB RAM.

]]>https://forum.mango-os.com/post/12144https://forum.mango-os.com/post/12144Mon, 25 Jul 2016 14:06:29 GMT<![CDATA[Reply to Mango hardware configuration on Tue, 19 Jul 2016 22:03:07 GMT]]>I used the free AWS ubuntu instance for a while to do some initial testing with Mango. It ran without issue with around 400 tags.

I am now using t2.large, and have no issues with around 2000 tags.

]]>https://forum.mango-os.com/post/12100https://forum.mango-os.com/post/12100Tue, 19 Jul 2016 22:03:07 GMT<![CDATA[Reply to Mango hardware configuration on Mon, 18 Jul 2016 22:13:45 GMT]]>That server will be more than enough. If using a Amazon server you could probably get away with their second smallest server but as Phillip mentions it depends a lot on how many users will be logging in and doing reports or viewing dashboards. For just the collection of data of 1000 data points every minute it doesn't take much to do that.

]]>https://forum.mango-os.com/post/12088https://forum.mango-os.com/post/12088Mon, 18 Jul 2016 22:13:45 GMT<![CDATA[Reply to Mango hardware configuration on Mon, 18 Jul 2016 19:18:56 GMT]]>@phildunlap said in Mango hardware configuration:

'm not sure what protocol you'll be performing this transport with, and you didn't specify the update rate either. Still, 10 sites * 100 points/site comes out to only 1000 points which is a pretty standard / low load, and I would think almost any multicore, 2GB+ RAM computer could handle it. If you perform a lot of analysis or reporting on these points at your central computer, that could require more power in the machine.
If you haven't seen it, a MangoES (http://infiniteautomation.com/index.php/hardware) is a really good value and could probably handle what you are describing, but if the whole setup will grow (more sites, more analysis in central, etc.) then you may wish to get an enterprise license so that you can transfer it if you need to upgrade your hardware at any point.

Hello phildunlap,
the protocol is Modbus/TCP and the controls are about every 60 seconds.
I saw MangoES but I prefer to start with a physical server because there is the opportunity to go in the next future on a cloud server.
Based on your experience could fit an Ubuntu server with 8GB of RAM, 1 CPU with 8 cores and 80GB HD.
Best Regards, Luigi

]]>https://forum.mango-os.com/post/12085https://forum.mango-os.com/post/12085Mon, 18 Jul 2016 19:18:56 GMT<![CDATA[Reply to Mango hardware configuration on Mon, 18 Jul 2016 18:48:45 GMT]]>Hi Luigi,

Welcome to the forum!

I'm not sure what protocol you'll be performing this transport with, and you didn't specify the update rate either. Still, 10 sites * 100 points/site comes out to only 1000 points which is a pretty standard / low load, and I would think almost any multicore, 2GB+ RAM computer could handle it. If you perform a lot of analysis or reporting on these points at your central computer, that could require more power in the machine.

If you haven't seen it, a MangoES (http://infiniteautomation.com/index.php/hardware) is a really good value and could probably handle what you are describing, but if the whole setup will grow (more sites, more analysis in central, etc.) then you may wish to get an enterprise license so that you can transfer it if you need to upgrade your hardware at any point.

]]>https://forum.mango-os.com/post/12084https://forum.mango-os.com/post/12084Mon, 18 Jul 2016 18:48:45 GMT