SOURCE: can I add additional drives to Raid 1 in order to expand storage?
I am not familiar with the specific model, but my guess is you can setup another separate logical vlume by using a RAID1 for the new disks. You will not be able to add the disks to the existing volume.
If there is no data, and the array supports it you can delete the existing volume and setup a RAID 5 instead, which will use all 4 disks, as long as they have the same capacity.
SOURCE: Loss of system access - no chance to backup data before fw-upgr.
I don't know if this problem is still open. I am amazed that this problem still existed in 2009, since I had this problem in 2007, and I handed Intel the solution at that time.
After escalating to the highest level of technical support at Intel, it turns out their answer was to ship my raid box to them and they would fix it, but there was no promise at all of whether or not my data would survive.
I wasn't about to lose two years of business data, as I wasn't in the mood to go looking for a new job.
Since I have a strong background in Linux, I decided to put the lead drive from the mirror set in a linux box and take a look at the partition. I learned that the SS4000-E keeps the root file system on the hard drive(!) instead of on the ROM that's in the box. Why? (A fully functional Linux kernel fits on a floppy disk.) I guess because Intel product development just wasn't up to creating a NAS. And of course, because it shaved the cost of the box by $5.
This is where Intel's "fix" comes in to play. The firmware that these units shipped with had a rather serious bug. If the unit shut down abnormally, it was about 90% guaranteed to corrupt the root file system on the drive. Intel recommends updating the firmware to make sure the problem isn't repeated, which unfortunately would erase all your data.
Intel was convinced there was no other solution.
If you have managed to read this far, here is the solution. The Linux kernel runs a chkdsk everything nth time the file system is booted. On the drive I stuck in a Linux box, I read the log to see how many reboots had been counted, and then config to see how many were required. I think the magic number was 15 more reboots. I put the drives back in the SS4000-E, rebooted 15 times, chkdsk ran and fixed the root file system, and the box went back online. I immediately flashed the firmware and the box worked perfectly for another three years.
Thank you, thank you. No danger of losing data!
This is a fundamental function of Linux. And no one at Intel knew this? Holy moly! I provided Intel with this solution. And they never shared it with anyone. Holy moly!
Intel makes some fantastic products. This time they were messing with things they just *did not understand*.
Know-it-all
486 views
Usually answered in minutes!
×