Did you try 7.0 kernel?
sudo modprobe tcp_bbr
sysctl net.ipv4.tcp_congestion_control
On 7.0 i get:
net.ipv4.tcp_available_congestion_control = reno cubic bbr
First!
Just kidding, tested with one physical atm, works okay.
Linux SPS2 7.0.0-1-rc6-pve #1 SMP PREEMPT_DYNAMIC PMX 7.0.0-1~rc6+1 (2026-03-30T09:17Z) x86_64 GNU/Linux
CEPH in Proxmox is the endgame, but as they said,if you really don't need that type of HA(or even real HA), you could go with storage replication. It works good, manual failover is fairly quick and with backups you're good.
Well,if you store Netflow that much(multi-TB usually), then it makes sense that db is in a multiple VM's ,and then maybe you can create a different storage on a same PBS with a different namespace. Atleast that is how i do it with NFA.
For my customers we are using rds farm + fslogic on CEPH tbh, not on local disks. Try changing cpu type to something instead of host. Also maybe screenshots of load and io of the VM,maybe you have something like I/O stall?
I use for some testing those nested snapshots(testing different versions of some app), and i haven't found a reliable way to migrate or backup those snapshots.