Hello all. (= I was wondering what the view/verdict of using Pandora FMS inside a OpenVZ container (for production)? I saw this:
It's not recommended to use Pandora FMS in virtual environments, since the disk access requirements are very high. In case you have to do so, it's strongly recommended to assign independent RAM, CPU and hard disks instead of using a part of the host's resources and to use SAN disks in these kinds of environments. It's strongly recommended to use Pandora FMS in virtual environments for testing, certification or development environments, but never in productive systems !
Source: http://wiki.pandorafms.com/index.php?title=Pandora:Documentation_en:Installing However, as OpenVZ is more of a 'jail'/chroot than a virtual machine.... does the same rules/views apply? Should I not be doing this (running Pandora FMS in a container, on a production box)? Im running <10 containers on a dedicated box: Dual Intel Xeon L5420, 16GB DDR2 ECC RAM, 500GB SATA HDD. Another question, if I wasn't to install FMS inside a container, but on the host itself, could I somehow use API calls via OpenVZ to monitor the containers, rather than deploying agents on each machine? E.g. I'm not sure if I’m on the right path, but it looks possible with VMware?: http://pandorafms.com/Producto/vmware-monitoring/en // http://wiki.pandorafms.com/index.php?title=Pandora:VMWare_ESX The host OS is Proxmox VE 3.3 (https://www.proxmox.com/) - latest at the time of posting. All "VMs" (aka jails, containers, or guests) are using OpenVZ.
