> A raid-1 configuration is defined as two Hard Drives (HD) having the > same computer software on each of them. If one HD fails the other will > continue to run and the computer will operate as if nothing had happened. > RAID-1 (I'd capitalize as it's an acronym - Redundant Array of Inexpensive Disks) It is not necessarily two hard drives with the same computer software. You can set up two partitions on one single drive and configure as RAID-1 (pretty useless as you take a performance hit and if the drive dies both partitions die - the mirrored copy included). You can have two partitions (each on a unique drive) that are set up as a software RAID-1. You can have 10 drives set up as RAID-1. So 5 drives are mirrored onto the 5 other drives. So RAID-1 is real time mirroring of a drive(s) or partition(s) onto another drive(s) partition(s) for redundancy (not backup). The other important point is that the drives should be the same size (not absolutely necessary but certainly recommended) if mirroring entire drives. After that you lost me. You say you are checking the partitions on the new HD, and then add a single partition on it. Then tell the person they must repeat this for /dev/sda1-7 (yet you just did it for 7, plus to a novice does that mean you repeat the same exact steps? What if the partitions exist? Don't exist? Are they all the same size - 30 gig?). You should start off the scenario for which this tutorial applies. For example: "This tutorial will walk you through setting up a software RAID-1 using an existing drive and a new drive where you wish to have the existing drive mirrored onto the new drive." If there is potential of data loss (and there always is when doing something like this) you need to warn people of that risk and advise that they proceed accordingly (i.e. backup if you have important data or can't afford to have to reinstall the original drive). You should also identify your intended audience and the necessary knowledge/skills recommended to use your tutorial. If you are not going to how to connect a drive and how to determine the device assignment, then that is a pre-requisite. If you are not going to explain how to delete a partition if you screw it up (typo for example and made it 20 gigs instead of 30) then you are assuming the person knows how to do that so it should also be noted as recommended knowledge. It's great that you are motivated to do up a guide for others. Many will surely benefit. But if not done properly you can cause much more grief than relief. There are existing guides out there. tldp.org has one dating 2004. Another is at http://www.linuxhomenetworking.com/wiki/index.php/Quick_HOWTO_:_Ch26_:_Linux_Software_RAID Yet another at http://www.enterprisestorageforum.com/technology/features/article.php/3526891 You may want to work through one of those and keep track of your experience (problems you encountered and how you solved them, tips/tricks) and then provide that to people. Tell them which guide to use, and provide your experience as an addendum to that guide. Setting up a software RAID is not a task that lends itself well to simply stating: type this, do that, then this without foreseeing potential problems an end user may have and warn of those and provide explanations on how to correct the problem should they encounter it. Something as simple as instructions on how to change ownship on a bunch of files across sub-folders can be dangerous. If the user executes the chown -R from the root by mistake and they possess the necessary rights they can really screw up their system. I give this example because I saw it happen in a class when chown -R was covered - one user changed ownership on every single file to himself. Luckily being a classroom where everybody had the same configuration I was able to write a script to grab ownership of files on a comparable system (except those belonging to the end user) and with another scripte change ownership back on all those files on his system. I agree with Tim. You can't take a topic like this and condense it to a page or two of "do this, then that, then this, then ...") unless you know your audience and their system configuration well enough to be able to do so (i.e. in a controlled, classroom environment that may be appropriate in some circumstances). Newbies will use such guides and follow them to the letter. But one typo, or one slight difference in setup and everything goes south pretty quick and you are cursed on. Jacques B.
