hi all,
First, thanks to all who contributed to Proxmox and this forum. I am new to Proxmox and just started using since last week. Learned a lot from this forum and I benefitted from your contribution. Still fine-tuning my system but I noticed a stability problem that I could not resolve. I seek your help.
I am running Proxmox on an Intel J4205 itx MB with 1 SATA SSD and 3x 14TB HHD. The system config is:
Other LXCs include the ARR stacks and jellyfin (with hardware transcoding). Everything works well except that, I noticed an instability problem when sabnzbd is "unpacking" a downloaded nzb file that was uploaded manually. Sometimes, the whole system would semi-freeze, and if I left it long enough, the whole proxmox would eventually crash.
It would appear that TrueNAS is the first to go; as all LXC has the NFS mounted using mount point, all LXC eventually crashes as well.
However, for any nzb file that was found and sent by the ARR stack, I have not had a single crash so far. Thus it would appear, and I hope I am not over-thinking it, that the crash only happens for a manually uploaded nzb file.
Previously I used Processors='host' for TrueNAS and had a lot of stability issues with heavy IO, changing the processor from host to kvm64 has reduced the instability problem during heavy IO but it still happens.
Can anyone suggest where to look for insights into this problem?
First, thanks to all who contributed to Proxmox and this forum. I am new to Proxmox and just started using since last week. Learned a lot from this forum and I benefitted from your contribution. Still fine-tuning my system but I noticed a stability problem that I could not resolve. I seek your help.
I am running Proxmox on an Intel J4205 itx MB with 1 SATA SSD and 3x 14TB HHD. The system config is:
- Asrock J4205-ITX MB with 16 (2x8GB) RAM (I know I should get more but 16GB ram stick is difficult to come by)
- SATA SSD serves as boot-disk and holds all VMs and LCXs,
- TrueNAS Scale as VM, 1st to boot (10GB ram, 2 CPU, cpu=kvm64, machine=q35, SeaBIOS),
- 3x14TB as a single ZFS pool, exported as NFS,
- Proxmox serves as a NFS client
- sabnzbd inside a LXC,
- boots after TrueNAS (180 seconds after TrueNAS to ensure the NFS is mounted properly)
- NFS mounted as a mount point via Proxmox (e.g. mp0: /mnt/NAS_NFS/,mp=/mnt/NAS_NFS in the config file)
- All download/temp files/directories are on the NFS and not inside the LXC,
Other LXCs include the ARR stacks and jellyfin (with hardware transcoding). Everything works well except that, I noticed an instability problem when sabnzbd is "unpacking" a downloaded nzb file that was uploaded manually. Sometimes, the whole system would semi-freeze, and if I left it long enough, the whole proxmox would eventually crash.
It would appear that TrueNAS is the first to go; as all LXC has the NFS mounted using mount point, all LXC eventually crashes as well.
However, for any nzb file that was found and sent by the ARR stack, I have not had a single crash so far. Thus it would appear, and I hope I am not over-thinking it, that the crash only happens for a manually uploaded nzb file.
Previously I used Processors='host' for TrueNAS and had a lot of stability issues with heavy IO, changing the processor from host to kvm64 has reduced the instability problem during heavy IO but it still happens.
Can anyone suggest where to look for insights into this problem?
Last edited: