I am trying to size my VM installation and I currently have it installed on a VM with 2GB RAM and 1 vCPU.
I notice when I run top, I quickly hit 80-90% us% and spike to about 10% on the wa%
I know I am on a 6+ year old Server 2016 server installation ( I have a new 2019 shipping )
I was thinking of giving it another vCPU but wanted to lean out the modules first as I think this should be best practices anyways.
I currently have a simple setup.
(3) trunks: contiguous US, International, e911
(3) extensions ( simple Ring Group to FXS ports on an Adtran; no conferencing, queueing/agents, etc )
(3) DIDs ( 1 of which is mostly used )
(2-4) concurrent calls.
I will be increasing this to the following:
(3) Trunks
(24) extensions ( 5-6 Ring Groups to FXS ports on an Adtran; no conferencing, queueing/agents, etc. )
(~7-10) concurrent calls. ( 24 max as the adtran only has 24 FXS ports )
I was thinking of disabling isymphony, zulu, and CID Superfecta.
Is there any other modules that should be set to disable to keep the background processes from firing?
Regardless of your system issues, you should always remove anything you are not using. Yes, my opinion, not everyones’. It means fewer things to update and fewer things to possibly break.
I run much larger installs on $5 Vultr instances.
I’m not sure what you are doing to see those kinds of loads for the described workload.
You want to remove isymohony (the service, not just the module) it uses 2gb of ram on its own. The Zulu daemon also uses a bit of resources. Superfecta uses nothing.