Minimum memory & CPU requirement for PMG (clamd hangs on 2GB systems)

[AlexHK]

I am currently evaluating PMG to replace a AWS EC2 based Windows 2019 mail server (MDaemon) running in a single AWS EC2 t2.small instance (1 vCPU 2GB RAM).

Today I installed PMG on a t2.small on top of the stock Debian 11 AMI and it would become unresponsive when clamd starts up. Changing the instance type to t2.medium (4GB) or t2.large (8GB) works though. It starts up fine on PVE7 with 1cpu 2GB (installed from ISO, not on top of Debian), also in VMWare Workstation, so I am not sure what is going on with AWS.

So now I am worried. MDaemon comes also with ClamAV and lots of other things, but still runs fine in the t2.small instance including all the M$ overhead.

Any idea how to debug this or find out why clamd startup will make the VM hang on AWS?

 

[AlexHK]

Hmm, the same happens when I install a 1 CPU 2GB container with the official PMG template (7.0.8) on a local PVE7 server.

clamd hangs on start and memory far exceeds the 2GB from the 'minimum requirements' page.

When I increased swap from 512MB to 2048MB it recovered.

And installing in a 1CPU 2GB VM on top of Debian 11 clamd fails altogether on startup, not just hang.

So it seems installed from ISO 2GB are fine, installed on top of Debian or running the official PMG template in a Debian LXC are NOT fine on 2GB of RAM.