Core Machine (Operating system/System info/Roon build number)
Debian GNU/Linux 10 (64-bit)
Linux roon 4.19.0-8-amd64 #1 SMP Debian 4.19.98-1 (2020-01-26) x86_64 GNU/Linux
I can’t get the Roon build number as I can’t get into the system long enough to figure it out, but I just downloaded it today
Intel® Xeon® CPU E5-2660 v2 @ 2.20GHz
Logical Processors: 40 [2cpu x 10c/20t]
Network Details (Including networking gear model/manufacturer and if on WiFi/Ethernet)
10 Gig Ethernet (fiber) via VMXNET3
Audio Devices (Specify what device you’re using and its connection type - USB/HDMI/etc.)
iPad Air (3rd Generation)
Description Of Issue
When importing albums to roon, the server becomes unstable consuming every ounce of being from it. But I’m not 100% sure that it’s only CPU… I have tried multiple settings to try and figure out what is causing the VM to lock up. But it’s not just the VM that locks up, it will lock up my entire host. The VM will throw erors like this " kernel:[ 646.768795] watchdog: BUG: soft lockup - CPU#1 stuck for 28s! [kworker/1:1H:180]".
I thought it was that hypervisor so I moved to to my gaming one with faster CPUs (Intel® Xeon® CPU E5-2690 v2 @ 3.00GHz [1cpu x 10c/20t]) but roon throws the same exact errors.
I have changed the virtual CPUs from 2 cores all the way up to 16 cores, but it doesn’t make a difference, same results of the roon VM locking up the entire host. I have also used the CPU limiter built into vmware and lowered the max frequency from 2.2Ghz down to 1.8Ghz in an attempt to not have it peg the system.
Storage from my NAS is over multiple 10G network links…
The storage for the VM is over a dedicated storage VLAN and is sitting on a SSD 3-way ZFS mirror, there is plenty of I/O and not hitting limits or await times on that array.
Secondary storage where my music library sits is presented via NFS to the VM (from my NAS) over a second (different) 10G network link. This pulled from a different ZFS pool that is a dual VDEV raidz array. This also doesn’t appear to be the bottleneck.
EDIT: Sad face, i think i figured it out, but only time will tell. Turns out that it might be the 10G Intel NICs that are causing the problem. My 3rd hypervisor has Mellanox 10G NICs and is no longer crashing. I am limiting the CPU still (down to 1.5Ghz) but overall the system seems way more responsive, and its been longer than 15 minutes while importing albums.