The GPU VM on the Powerspec box runs a 27B model that keeps its experts resident in RAM, so I wanted more RAM for it. The host has 64 GB, the VM had 50, and I figured I had room to grow. I bumped the VM to 56 GB and the host OOM'd. Twice.
This is the story of why, and it's mostly a story about assumptions I didn't check.
The first run of the health watchdog reported pbs-backup-local as inactive, a CRITICAL on a backup path that was perfectly fine. The storage was healthy the whole time. What the watchdog actually caught was a momentary auth handshake, and it treated one bad sample as ground truth.
That false alarm is why the watchdog works the way it does now.
The voice assistant on the HA box has three legs: speech-to-text on the GPU box, an LLM brain, and text-to-speech back out. For weeks, a full pipeline run over the websocket died in about 30 milliseconds. The whisper bridge log showed a client connecting and immediately disconnecting with zero audio. Read that as "HA can't reach the bridge" and you'll chase it for days.
It wasn't HA. It wasn't the GPU. It was my own test client, and the whole thing came down to one byte.
The PBS datastore on the external drive was shared into the VM over a PVE virtiofs mount. Backups failed roughly half the time, always with the same error, os error 23 — ENFILE, too many open files. Half the 03:00 UTC vzdump window was going to waste, and the failure looked random.
It wasn't random, and the fix wasn't a better mount.
I ran this for a while with two agents on the same engine and both of them felt sluggish, and I couldn't work out why. The engine reported good numbers. The experience didn't match.
The answer is the prefix cache, and it's structural — not a tuning knob.
In this post I'm going to discuss project and user agents for each of the platforms as well as native and third party applications that allow remote access to your agents.