Monitoring hardware / cloud

[wertert2]

Hi All

I've been asked by a friend for some advice. He wants to setup a small data centre and host a number of servers and a private cloud appliance. I think he plans to re-sell these resources onto various customers.

So the question he's asked me revolves around how can he monitor this equipment ? Years ago I used a couple of products by Heroix to monitor hardware via snmp traps / log files / console connections etc. Various alerts were run through a ruleset and would result in an email or sms to the appropriate operator/owner.

I was wondering if anyone out there in IT land had any recent up-to-date experience in this area ?

 

[BongoHunter]

For monitoring the actual hardware you'd be looking at your server vendors provided tool, such as Dell OME or HPE Oneview/SIM - these can be hooked up to your trouble ticket solution of choice or just send emails to an inbox

At the Virtualisation layer - use VCenter for you monitoring and alerting if going VMWare, if kvm/qemu based then look towards an open source tools.

If it's a full opensource stack or you want the cheapest option then most people will lean toward Nagios, or one of its derivitives like Icinga or CheckMK, Zabbix is also popular, there are loads of closed source tools in this space to like Solarwinds etc.

Avoid anything by MicroFocus (they have the former HPE tools) as they will probably going bust soon.

If I had a greenfield site with servers, network, apps and virtualisation to monitor right now I would be looking to do it entirely with Prometheus with its AlertManager addon for everything - it's modern, scalable, free and well maintained, and you can integrate other tools with it very easily - hopefully it will be the death of Nagios and it's kin.

 

[wertert2]

For monitoring the actual hardware you'd be looking at your server vendors provided tool, such as Dell OME or HPE Oneview/SIM - these can be hooked up to your trouble ticket solution of choice or just send emails to an inbox

At the Virtualisation layer - use VCenter for you monitoring and alerting if going VMWare, if kvm/qemu based then look towards an open source tools.

If it's a full opensource stack or you want the cheapest option then most people will lean toward Nagios, or one of its derivitives like Icinga or CheckMK, Zabbix is also popular, there are loads of closed source tools in this space to like Solarwinds etc.

Avoid anything by MicroFocus (they have the former HPE tools) as they will probably going bust soon.

If I had a greenfield site with servers, network, apps and virtualisation to monitor right now I would be looking to do it entirely with Prometheus with its AlertManager addon for everything - it's modern, scalable, free and well maintained, and you can integrate other tools with it very easily - hopefully it will be the death of Nagios and it's kin.

Great - thanks for the info - loads of really useful pointers. I'll do some research.