Venus OS v3.81 / v3.79: Hourly VE.Bus / BMS communication drop during internet outage (Cerbo GX Off-

Bit of a niche one this, but wondering if anyone else has clocked this on their setup.

Running a MultiPlus-II 3000 on the narrowboat with a Cerbo GX and a CAN-bus BMS (Fogstar Drift 200Ah lithiums). DVCC enabled, no ESS — pure off-grid.

Started noticing the inverter throwing brief faults roughly on the hour. Nothing catastrophic, but the system was clearly having a wobble — charger dropping out for 30-60 seconds then recovering. Took me an embarrassingly long time to correlate it with the marina WiFi being patchy. Once I killed the Cerbo's internet connection entirely, the hourly blips stopped dead.

Dug around and it looks like this is a known gremlin in Venus OS v3.79/v3.81 — something to do with how the Cerbo handles VE.Bus or BMS comms when the internet drops. The communication stack seems to hiccup on whatever scheduled check it's doing, and that's enough to momentarily confuse the BMS handshaking.

A few questions for the room:

• Anyone else seen this on a narrowboat or static caravan setup where the internet connection is genuinely unreliable?
• Is the fix simply staying on v3.73 for now, or has someone found a workaround that doesn't involve rolling back?
• Does running the Cerbo in offline mode fully resolve it, or just reduce frequency?

Would be good to know if Victron have acknowledged this officially anywhere — I've seen forum chatter but nothing concrete from their dev team. My system runs unattended for days sometimes, so hourly comms drops aren't exactly comforting when you're

@HeatherSoul yes, seen this exact behaviour on my static caravan setup. Cerbo GX running v3.79, MultiPlus-II 5000, Fogstar Drift 206Ah via CAN.

The drop happens on the hour, every hour, lasts roughly 8-12 seconds. Logs show it coincides with whatever NTP sync or cloud heartbeat the Cerbo's attempting when it can't reach the Victron servers.

Workaround that's mostly stabilised mine:

• Set a local NTP server on your router
• Disable "Remote Support" in the Cerbo settings if you don't need it
• Set VEConfigure keep-alive interval shorter via SSH if you're comfortable in the terminal

Not a permanent fix — it's clearly a firmware bug where the comms stack blocks during failed network requests. Victron community forum has an open ticket. Worth adding your details there with exact firmware string so they can reproduce it properly.

Seen something similar on my shepherd's hut setup — Cerbo GX on v3.79, Fogstar Drift 200Ah, MultiPlus-II 3000. The hourly drop is very consistent, almost like clockwork, which makes me think it's a scheduled task or keepalive timeout rather than random packet loss.

Worth checking your /data/log/dbus-systemcalc-py/ logs during the outage window — mine showed a VE.Bus resync attempt right before the drop.

@FormerMechanic14 did you notice whether the BMS comms drop before or after the VE.Bus blip? On mine it's BMS first, which makes me suspect the CAN watchdog is the trigger rather than the MultiPlus side.

Raised it on the Victron community forum too but no official response yet — might be worth us both logging it as a bug report with timestamps attached.

Interesting one – I can add another data point here. Same pattern on my off-grid cabin setup, Cerbo GX on v3.81 this time rather than v3.79, so it's persisted across that update at least. What I've noticed is the drops seem to correlate almost exactly with the Cerbo attempting its hourly VRM sync and timing out when there's no internet. Almost like it's blocking the local BMS comms briefly while it waits for a response from the VRM servers.

Worth checking your VRM connection interval settings if you haven't already – I reduced mine and the drops became less disruptive, though didn't eliminate them entirely. Also, have you both logged a ticket with Victron directly? This feels like something they'd want reproducible data on. The more setups reporting it the better, really.

Ran into something eerily similar on my motorhome last winter — Cerbo GX on v3.79, MultiPlus-II 3000, Fogstar Drift 200Ah on CAN-bus. Was parked up in the middle of nowhere with no mobile signal, and the hourly drops were driving me mad trying to diagnose a fault that wasn't really there.

What caught my attention: the drops seemed to correlate almost exactly with the VRM portal retry attempts. Like the Cerbo was briefly choking on failed outbound connections during that cycle.

@KenCrane — interesting that you're seeing it on v3.81 too. I'd half-hoped Victron had quietly patched something between versions.

Have any of you tried forcing the VRM portal connection to disabled entirely rather than just letting it fail? I'm curious whether removing that retry loop from the equation changes the drop frequency at all.