Support and News Forum

I'm following the instructions for installing opnsense from the current DVD image into a qcow disk image and it's unbelievably slow. As in, after 18 hours it's still only 57% of the way through copying files from the DVD image to the hard disk. Is this some problem with setup, or is that simply how long it takes?

I'm running Eve-NG Pro 4.01-16. This is the only node running on it.
Eve is installed as a VM, running on an ESXi 6.5 host.
The host is a ProLiant DL385p Gen8, with two AMD 6344 12 core CPUs and 256 Gb RAM. It's running a couple of other VMs but nothing too taxing.
The Eve VM has 8 cores and 24Gb RAM, but with just a single node running it's not coming close to CPU or RAM limits.
Initially it was running on local RAID of HDDs, but I moved the VM onto shared storage running on an all-flash TrueNAS core box, where it's currently the only VM on the storage so not having to compete with other use, but that made no difference.

I'm assuming the issue is with qemu and disk IO? Is there anything I can change on the system or in the opnsense definition file to make it behave better?
Other images seem to work OK but they tend to be much smaller.

NAS VM will be only worse
performance of pfsense or opnsense depends of CPU and ram assigned per node..

Other than that opnsense is maxing out one core, it's not starved for CPU or RAM - it's purely disk IO that seems to be the bottleneck, and purely within qemu. I'm guessing it's because it's writing a lot of data via COW into the overlay. I'll just be patient ( I killed off the node) and then see how it performs later. I think I'll try increasing the resources in the template while it's doing the install from CD image to see if giving it more RAM to use as a cache helps.