Broken Solar Yield Graph in Venus OS 3.x

Been scratching my head over this one for a couple of weeks now and wondering if anyone else has run into the same wall.

Since updating to Venus OS 3.x on my Cerbo GX, the Solar Yield graph has basically stopped working properly. I've got three Victron MPPT controllers running off the roof array on the van — a 100/30, a 150/35, and a smaller 75/15 for the cab panel — and when I navigate into Solar Yield and drill down into any individual controller's history, I'm only getting a day or two of data before it just... stops. The graph renders but it's mostly empty space.

The full fleet overview looks fine at a glance, but the moment you want to actually interrogate individual controller performance over a week or a month, it's useless. I rely on this data to understand how the different panels are performing through the seasons — essential when you're parked up somewhere shaded and trying to work out whether the problem is cloud, angle, or a dodgy connection.

Worth noting I'm running the New UI that shipped with 3.75. Not sure if reverting to the classic interface would restore the history or if the underlying data is being lost entirely.

A few specific questions:

• Is this a known bug in 3.x that Victron are working on?
• Does reverting to the old UI recover the historical data?
• Has anyone found a workaround — VRM portal perhaps, or pulling data another way?

Running a Fogstar 280Ah LiFePO4 bank if that's at all relevant to the setup context. Cheers for any pointers in the right direction.

@CoastalCamper yeah, classic Venus OS 3.x headache this one. The yield graph breakage is usually down to the VRM portal losing sync with the local MQTT data after the update — particularly if your Cerbo's system time drifted during the upgrade process.

Few things worth checking:

• Settings → System setup → Time zone — confirm it's set correctly (Europe/London catches people out after DST)
• SSH in and run date to verify actual system time matches reality
• Check whether your MPPT is on the correct VE.Direct port — 3.x reshuffled some device enumeration

On my setup (Victron SmartSolar 150/35 feeding into a Cerbo) I had to do a full VRM authorisation token reset after upgrading, which magically brought the graphs back to life. Settings → VRM online portal → Reauthorise.

What MPPT are you running and how's it connected?

Had this exact issue on my garden office setup after updating the Cerbo. Worth checking whether your timezone settings are correct on the device — mine had somehow reset to UTC after the update, which was throwing the daily yield totals completely out of whack and making the graph look broken even though the underlying data was fine.

Go to Settings → General → Timezone and make sure it matches your actual location. VRM pulls that timestamp data to build the graph, so if the device thinks it's a different day it'll mess up the bar chart boundaries.

Also double-check under Settings → VRM Online Portal that logging is still enabled — the update can occasionally toggle that off silently.

If neither of those fixes it, try a full power cycle of the Cerbo rather than just a reboot. Sorted mine.

Has anyone tried rolling back to 2.x to confirm it's definitely the Venus OS update causing it rather than something else? I updated my Cerbo a few months back and noticed some oddities with historical data but couldn't be certain whether it was the update or a VRM portal sync issue.

Also — does the broken yield data affect the battery state of charge history as well, or is it purely the solar yield graph that's misbehaving? Trying to figure out if it's a display/reporting layer problem or something deeper in how Venus OS 3.x is logging the data from the MPPT.

My setup is a Victron SmartSolar 100/30 feeding into a Fogstar 200Ah, so curious whether MPPT firmware version plays any role here too.

@George1972 rolled back on my shepherds hut setup specifically to test this — 2.94 restored the yield graph immediately, confirming it's definitely a Venus OS 3.x regression rather than hardware or VRM account issues.

That said, I wouldn't stay on 2.94 long-term. The fix I've settled on is forcing the MPPT history reset via the Cerbo's console, then letting VRM repopulate from scratch. Took about 48 hours before the graph looked sane again.

Worth noting: if you're running multiple MPPTs (I've got two Victron 100/30s feeding the hut), the graph corruption seems worse — each controller's yield appears to overwrite the other's daily total rather than aggregate properly. Single-MPPT installs may see a milder version of the bug.

Check the Victron community forum too; there's a pinned thread tracking which 3.x point release is expected to patch this.