I have a problem that is in common with a lot of IT Pros. I have a lot of data that I value and I want to keep it save. ALL forms of storage are flawless – disk die, SSDs, die, tapes die, tape drives die, CDs die, floppy disks die. So what to do?
IN my case, my data is my collection of Grateful Dead, Jerry Garcia and a bunch of other music. The Dead stash takes up around 1.6 TB, with the Jerry and other music weighing in at around 1 TB. That’s a lot of music (and with around 1900 Dead shows, one of the larger around!). I do not want to lose that. The music has some interesting characteristics that may or may not be shared with other data storage scenarios. With music, each show or album is made up of multiple songs. As most of what I collect is lossless with two formats. SHN (or Shorten) and FLAC (Free Lossless Audio Codec). Typical GD shows range from 750MB to 1GB or so. Typical ‘song’ sizes are anywhere from 30/40MB to 120MB and sometimes more (and sometimes less). Once in my hands, these files are very much read-only!
This data profile is different from many simple file servers with lots of very small documents. Most of which don’t change very often.
The profile is also different from database servers where there are a few VERY big files (containing lots of bits of occasionally changing data). In some applications, like LYNC, you have several databases that need to be backed (and restored in case of failure).
Given the varied nature of data usage, there are inevitably different approaches for backup.
For my music collection, I have 4 big capacity USB-disks. One pair holds the master and are tethered to my main workstation where I update these as I can. The other, backup, pair are on the other end of my network. I have two scripts that use Robocopy to sync primary to backup. Over the years, I’ve been bitten badly by disks dying. So what I do is to wait till one of the disks begins to fail, and I replace it with a new drive. And even if it looks OK, I rotate every couple of years.
Recently, the 2tb disks I used to for the GD stuff were filling up, so time to move up format, and I’ve not got my collection on new 4TB disks. As per the title, it took a while to do this copy. Here’s the Robocopy Log:
So, to copy my entire GD collection from one USB drive to another took a mere 30 hours. Not bad when you consider I’ve been collecting for 30 years.