Hi, I’ve installed Rockstor on a USB 3.0 stick in order to use all my disks for data storage. What I want to achieve is to have a RAID 1 for this USB stick. I’ve learned that Rockstor does not support this (yet), so I’m looking for a workaround.
The next thing I can think of, is to clone the USB stick to another one. Is there any software out there that can do the job online? Many tools are either offline or they write to iso files.
Thanks a lot,
Gregory
Get Clonezilla, I believe there’s also radymade bootcd available.
@greggy Welcome to the Rockstor community. There are issues with linux booting multi device btrfs as root, @Spectre694 is probably the furthest along with experimenting in this area on Rockstor and has referenced the various upstream blocker issues in the following forum thread:-
Install to full disk BTRFS
There is a workable solution for advanced users that employs mdraid to mirror the system disk’s partitions independantly, this is documented in the official docs over at:-
Mirroring Rockstor OS using Linux Raid
If you just wish to take an (offline) image backup of a Rockstor USB install then this could be done using the built in linux dd command, though it would have to be done on another linux install, not the Rockstor install itself, although you could use the recovery mode of the installer I imagine. Ie dd from the installed Rockstor USB system device to a file and then reverse the process to another similarly sized USB key. Obviously you couldn’t use them both at the same time but it would serve as a manual fail over for a key or config failure. But this obviously doesn’t meet your ‘online’ criteria.
The dd command in linux and OSX for that matter are documented on the Quckstart doc page in the Command line options section.
Please note however that until Rockstor moves to a >= 4.5 kernel there are known issues with boot from usb devices with the kernel versions > 4.2.5-1.el7 and less than 4.5.0-0.rc1.el7 such as is standard in Rockstor at the moment.
We have an issue open on this and since it has now been fixed in upstream elrepo kernel versions this should be sorted as soon as we move to a new enough kernel as standard. This info is thanks to @roweryan here in the forum and over at the issue where he references the upstream issue and it’s resolution.
Hope that helps.