Establishing a ControlUp Infrastructure Across Two Sites

A user sought advice on building a ControlUp infrastructure to monitor two geographically distant sites, separated by 2500 km/1500 miles. Experts suggested the user leverage the latest version (8.8) of the ControlUp Console and Optimization Pack, and place all the necessary components, including the Solve and Insight Pack, in the larger site once it is built and ready. It was also suggested that it might be possible to establish a temporary network in Site A to help facilitate easier movement of components. The ControlUp Academy can be found at https://cuacademy.controlup.com/.

Read the entire ‘Establishing a ControlUp Infrastructure Across Two Sites’ thread below:

Hi guys,

I would like to have some advices to build a main Control Up infra that must monitor 2 different geographical places for my customer. They are far away of 2500km / 1500 miles !

The place A is built and monitor about 300 VDI (small site). The place B will be built soon but will be bigger with 1200 VDI.

I planned to deploy 1 management monitor and 3 monitoring monitors on place B. They will be linked and managed by my place A where the COP server, RT Console, Solve and Insights are already deployed. Of course, I will configure a “Monitor site place B” in my Control Up configuration to gather the place B monitors together.

But I’m wondering if it won’t consume the network bandwidth by sending all the metrics from place B to place A.

Furthermore, can I deploy some IOP data nodes on place B, configure the Splunk config file of the place B monitors to send their metrics to these specific IOP data nodes instead of sending them to the IOP cluster ? How the IOP cluster will behave with IOP data nodes on place A and on place B distant from 2500km ?

What will be my bandwidth network consumption ?

Migrating existing infrastructure from place A to place B is currently not wished.

Thank you for your helpful feedback experience, adivces, analysis.

It’s a big challenge for me and I want to prove the Control Up’s efficiency even on such particular environments. Thank you.

What do you think @member @member ? I am quite sure the RTDX components will manage the latency easily, as it’s just some communication to the COP services and monitor management traffic over 40706. But what about IOP? Can we have IOP “sites”? Or better to just place the entire IOP cluster in site B, and send the rather shall Monitor activity files from site A directly to IOP nodes in site B?

And @member , make sure you use latest COP 8.8, to simplify the networking requirements between the sites

You send iop data to the master not the data nodes so that won’t help. I agree that the latency is not an issue. With the improvements in 8.8 our data consumption between monitors and agents has gone down because of the process reduction work so I would be tempted to place all monitors in site A. I have customers where there’s a good chunk of latency between monitors and agents in Asia without issues. If you have monitors in the second site you not only have traffic to iop but also to solve doubling it.

@member I’m using the 8.7 version right now but I plan to deploy at the beginning of december the 8.8 version. I’m just a little afraid about the IOP nodes issue discussed in a Matthjs’s post just above.

Thank you @member: I also though about the SOP traffic for the RT. So that’s not a good idea to split the monitors on both places in that case. I have to check if our core infra has enough capacity to onboard these resources.

And what do you think if the SOP is moved to the place B also ? :thinking_face:

Well, thank you very much for your help, but don’t forget we are on saturday so enjoy your weekend ! Not Prodigy, but I will DJ tonight for a party 😉

if you place sop in the other side the traffic from site a will need to go there but I do agree that it makes the most sense to place our components in the largest site. (missed that B will become the bigger one at some point as I was on mobile earlier)

is it possible to have a site B network temporarily in site A so you can ore easily move the components later?

Continue reading and comment on the thread ‘Establishing a ControlUp Infrastructure Across Two Sites’.  Not a member? Join Here!