Hello, I haven’t been using my server for some time, and it wasn’t happy after the motherboard battery failed, it wouldn’t even show the HP logo and option to enter the Bios settings on power up.
Running Rockstor 4.1.0.0 on OpenSuse Leap Linux: 5.3.18-59.37
Have two RAID 1 pools, each with two disks
Brief description of the problem
HP Microserver Gen7 N54L wouldn’t booting at all after being powered off for a few weeks/months. I removed the four data disks, replaced the motherboard battery and reset the CMOS. I don’t think I let it boot up past entering the bios before putting the disks back in.
The web gui Storage/Pools window is now showing both pools as unmounted, there are two RAID1 pools, each having two of the four physical disks. The four disks are listed with the word “detached” at the beginning of their names.
The screen plugged into the HP server shows a final message 9.078620] FAT-fs (sda2): Volume was not properly unmounted. Some data may (be?) corrupt. Please run fsck.I’m not a Linux expert but don’t think this is referring to any of the four data disks. This server console window is not responding to anything from the keyboard. I can get to the system shell from the WebGUI.
Could I delete the pool and create a new one without losing data on the disk, or am I needing to do some recovery work on the disks first? I’ve been looking at similar threads on the forum, and have tried the btrfs dev stats -z /mnt2/Pool1command and btrfs dev stats /mnt2/Pool1 but they didn’t change anything. Any advice would be appreciated, I’m fairly certain most of the data has already been copied off, but I’d like to be able to check what is on this server.
It’s strange that you can’t get to the service console/doesn’t respond to anything from the keyboard. Which model of the HP Microserver are you running?
From the command line (system shell from webui or see whether you can ssh into the server using PuTTY or similar) can you perform a lsblk to ensure that the devices are visible to the operating system, as well as a btrfs fi show?
It might be that you have to mount the pools as degraded (manually) to see them.
Of course, there could always be a hardware connectivity problem leading to the detached status. If you pulled the disks just make sure the connections are tight again.
You can also take a look at tail -n200 /opt/rockstor/var/log/rockstor.log and journalctl -xe to see whether any messages stick out related to the mounting.
I kept reading that when you replace the CMOS battery on many of the Microservers, you should leave the new one out for 10-20 minutes, and that will automagically fix all boot up issues . Well, things like the keyboard not being recognized correctly, don’t know about the drive/pool issue you’re experiencing.
Hello Dan - I tried all your suggestions, starting with re-checking the cables, as it seemed like the server wasn’t seeing the 4 disks. I’m not a Linux expert, and initially thought the lsblk was showing 4 disks, but it was in fact showing 4 partitions on the boot (non data) disk. So re-checked the cabled and found one cable wasn’t quite properly inserted on the motherboard. Re-seated it, powered on the server, and the pools are now showing as available and I can connect to the Samba shares from my Windows desktop. I’ll blame the limited light and lack of space in these microservers!!
The console window still doesn’t respond to the keyboard, and it still leaves a load of weird characters on the screen after booting, but it started doing that after my last upgrade over a year ago, so I won’t worry about it for now. At least I can now verify everything is backed up, and backup anything I might have missed earlier. Then I can look at rebuilding it with a newer version of Rockstor, and replacing the disks with a couple of newer larger ones.
Thank you again for your suggestions and help - sorry it turned out to be due to me rushing to reassemble it and get it online again!
. Well, things like the keyboard not being recognized correctly, don’t know about the drive/pool issue you’re experiencing.