Okay. So far, we’ve got the motherboard, the processor, the memory, and the power supply. Along the way, I mentioned an after-market fan for the CPU. In this post, I’ll go over the other elements I put inside the case.
Of course, one of the things you need is a hard disk. In my case, I decided to get two identical hard disks. Alas, this is one of the places for a hard-learned lesson. Hard disks can use different kinds of technology to interface with the motherboard and the computer. You choose which one in the system BIOS setting. In my case, the choices were RAID, IDE, or ACHI. These are pretty standard. I already knew I needed to use RAID for the hard disks, since it’s required for the Intel RapidStorage Technology (RST for short) I mentioned back in part two of this series.
Now, there are several flavors of RAID. For the purposes here, I originally thought I wanted RAID1, which causes the two disks to “mirror” each other: everything is written to both disks. The idea is that you then have an automatic backup. If one disk fails, just pull it out, order a new one, and your system continues running just fine while you wait for the replacement to arrive. It turns out there are reasons why this isn’t an optimal backup strategy and you need to do ordinary backups anyway, but that was my original plan.
The actual cabling that you use for your drives–whether they are hard drives, solid state drives, optical drives–is something called a SATA cable. I can plug up to six SATA devices into my motherboard. I’d planned to use two for hard disks, one for a solid state drive, and one for my optical drive, leaving two left over for expansion.
So, I needed to have the two hard drives configured as RAID. The RapidStorage technology requires a solid state drive, also configured as part of the RAID array. No problem, thought I. But, of the six SATA ports, the Intel H77 chip controlled four and a third-party chip controlled two. In order to use RST, I’d have to put the hard disks and the solid state drive on the ports controlled by the Intel H77 chip. However, these ports operated at different speeds: two at 6GBs and two at 3GBs. No matter which device I plugged into which port, I couldn’t get a system that would (a) boot; (b) put the two hard drives on a RAID1 array; and (c) enable the RST technology. I concluded that the the three drives would not only have to use the same controller chip, but also all be at the same speed. This bit of experimentation cost me about half a day.
What I finally did was use a single RAID array for one of the two hard disks and the solid state drive, both attached to ports controlled by the H77 chip. Then I attached the second hard drive to a 6GBs SATA port controlled by the third-party chip, and not configured as part of my RAID array. Now everything booted, I had RST running, and the system was stable.
The final configuration of my drives is then one 2TB drive that has all my files on it, one 64GB solid state drive that’s the “swap drive” used by the RST, and a second 2TB drive that has only my weekly backups. I can still recover my system should the main drive fail since I’ve got backups, and all the data drives are connected to high-speed SATA ports.
For the solid state drive, I selected a 64GB version based mostly on price. To use RST, you need a minimum of 16GB, but the larger drive you’ve got the better. On later builds, I chose a more expensive and faster SSD, but truthfully I don’t see any difference in speed.
I wanted to be able to play the occasional BlueRay disk on my machine, so I knew that I wanted an optical drive with that capacity. With a little bit of research, it turned out to be not terribly more expensive to purchase one that would also write BlueRay disks. Another great feature of new disk drives is something called LightScribe technology. With a suitable disk, you can flip it over in the disk tray and the laser will write a label for you on the disk. There’s no more need to have sloppy and–in my case–illegible labels on your backup disks. How great is that?
I don’t know about you, but I work from multiple computers. I’ve got one in the living room so I can work while watching TV. It’s useful other ways, too, like discovering that really is Ellen Barkin in “The New Normal.” Then there’s the netbook on my nightstand that annoys my partner at 3AM when I get this fantastic idea I just have to write down. Of course, sometimes I need to work at home on work-related things. That could be on either one of the laptops mentioned above, or on my desktop system. I used to rely on VPN tunneling for the work-related files and file sharing at home, but neither of those work when I’m traveling, say, for the Higher Learning Commission or visiting family. Of course, there are the ubiquitous thumb drives, but they pose security problems if you lose them. Nowadays, I mostly use Dropbox for non-work-related file sharing. It’s secure, easy, and means I can work on documents anywhere I want.
Even with the above technologies, there are still instances where I sometimes want to be able to read or write to SD cards. For example, I’ve got one of those electronic photo albums on my desk that’s loaded up with a gajillion pictures of my grandkids. It takes an SD card and the only way to load it up with pictures is by plugging it into my computer. Thus, I wanted a card reader. There are lots inexpensive choices out there. I selected one that plugged into a fast USB 3.0 port on the motherboard and gave me an additional USB port on the front of my system.
I used an existing monitor for my system. It had VGA, DVI, and HDMI inputs. It also has speakers, which are more than sufficient for my limited purposes. By using the HDMI input, I minimize the cables that run from my system to my monitor.
I’ll finish this series with a discussion of cases.