Dealing with the “No Valid NVDATA Image” error on a Fujitsu PRAID EP400i (LSI3108)
If you ever find yourself staring at a RAID controller that suddenly refuses to show your drives, you’re not alone. I recently ran into a frustrating issue while trying to switch my Fujitsu PRAID EP400i (based on the LSI3108 chipset) from RAID mode to JBOD mode, hoping to use ZFS with my drives. Instead, I ended up soft-bricking the controller with the dreaded “no valid NVDATA image” error.
What Happened?
My setup involved a Fujitsu PRIMERGY RX2540 M2 server and a PRAID EP400i controller. Everything was humming along fine with a RAID6 configuration until I decided to switch the controller’s personality mode to JBOD. That’s when trouble started. After rebooting, my server greeted me with the message:
“Firmware did not find valid NVDATA image. Please program valid NVDATA image and restart your system.”
The drives stopped showing up entirely, and although the controller wasn’t totally dead—it let me into the MegaRAID BIOS Config Utility (MRBCU) and the StorCLI preboot CLI—none of the drives or virtual disks appeared.
Understanding the “No Valid NVDATA Image” Error
NVDATA (Non-Volatile Data) is crucial for your RAID controller’s operation—it stores configuration and state information. Without a valid NVDATA image, the controller can’t properly identify or manage your drives. This became a blocker for me.
What I Tried to Fix It
- Factory Resets and Resets: I dumped MRBCU’s factory defaults and also tried StorCLI commands like
/c0 set factory defaults,flasherase,flushcache,restart, and even deleting NVRAM. None brought back the drives. - Firmware Flashing: Using Ubuntu Live USB and StorCLI, I flashed both official Fujitsu firmware versions available — 24.21.0-0151 and 24.21.0-0076. The firmware updated successfully, but no luck with the drives.
-
Megarec Tool: I tried Megarec on FreeDOS to detect the controller, but it couldn’t find it at all.
The Current State
- I can access the MRBCU and StorCLI tools.
- MRBCU shows “Unknown” under VD Management and no drives under PD Management.
- Controller status reported as “Optimal” in StorCLI, but Cachevault and Battery Backup Unit (BBU) info is missing even though BBU hardware is present.
- Changing controller properties leads to operation errors.
What If You Need to Force HBA / IT Mode?
Flashing the controller into IT mode (which forces it to behave as a Host Bus Adapter) is often suggested when RAID functions won’t come back. It’s a way to bypass RAID management and let the controller behave like a simple drive interface.
However, doing this requires careful handling of firmware and utilities specific to your LSI3108-based card. This step isn’t for the faint of heart, but if your priority is just to get your drives recognized rather than RAID functionality, it’s worth considering.
Lessons Learned and Tips
- Always back up your controller’s configuration if possible before changing modes.
- Be cautious switching controller personalities—especially with older or less common hardware.
- Keep records of original firmware versions; sometimes older versions aren’t easily found but might be necessary.
- Check official firmware and utility tools from Fujitsu and Broadcom (who own LSI tech).
Helpful Resources
Final Thoughts
Dealing with RAID controller issues like the “no valid NVDATA image” error can be maddening, but don’t lose hope. Having access to utilities like MRBCU and StorCLI helps a lot. If you’re stuck, reaching out to community forums or tech support can provide new insights.
If you’re comfortable experimenting, flashing to IT mode might be your last resort to breathe life back into the card—especially when data recovery isn’t a concern.
If this has happened to you or you’re exploring similar upgrades, hopefully, my experience helps you avoid the same pitfalls or guides you back if you’re caught in the same spot. Hardware quirks can be tough, but persistence and a bit of digging can pay off.
Good luck, and here’s to hoping your RAID controller troubles stay in the past!