Hard Drive Upgrade

I finally put too many photos and videos on my laptop hard drive and I had to upgrade it. My original hard drive was a 640 GB drive and there are now 1 TB laptop drives available that have 8 GB of flash cache. Not bad for $129. I was able to move my Windows install to the new drive using my weekly backup. The upgrade went smoothly, with just one annoying glitch.


The travails of setting up my weekly backup were discussed previously in Backin’ Up is Hard to Do.

There is something marvelous about going from this:

image

to this:

image

Plus, since my laptop has a second drive bay I can actually have both drives installed simultaneously, giving me room to store a gajillion libraries of congress.

My weekly backup uses Windows Backup to do a file-by-file backup of all of my data files, and an image backup. The image backup is handy because you can use it to restore a complete bootable image to a different drive. After doing my regular backup (being careful not to use the machine during the backup, since any changes made then would be lost) I followed these instructions. This basically involves creating a recovery disk, putting in the new drive, then booting from the recovery disk and restoring to the new drive.

The backup took about twelve hours and the restore took about five. Be patient.

After doing the recovery and booting from the new drive you have to expand the partition size. I couldn’t find the “Create and format hard disk partitions” tool discussed in the instructions above. Instead I ran Computer Management as administrator and selected Disk Management. It was straightforward to go from this:

image

to this:

image

I found it vaguely impressive that no reboot was required during this stage.

Metric is the future, not binary

Windows insists on telling me that my 1 TB drive has only 916 GB. That is because Windows still thinks that consumers want to see file sizes and drive sizes described in terms of base-two KB, MB, GB, and TB. The base-two sizes make sense for developers who are reasoning about cache lines, cache sizes, and MMU pages, but they have no meaning in the world of file sizes and drive sizes. It is long past time for Microsoft to use base-ten sizes. There needs to be extraordinary reasons to justify using the ISO prefixes in a non-standard way, and there are no such extraordinary reasons. I shouldn’t have to multiply all of the GB sizes in the Disk Management UI by 1.0737 in order to verify that I’ve got the drive that I paid for (1.0486 for MB, 1.024 for KB, and 1.1000 for TB).

During a recent install of a Microsoft product I was told that I need 4,732 MB free. Since that is presumably a binary MB that is 4.621 GB, or 4,845,568 KB, or 4,961,861,632 bytes. Confusing enough? Put another way, in what world does it make sense for 1,020 MB to be equal to .996 GB? There’s no reason for binary in this context. None. Apple’s file manager has made the switch to base ten, it’s time for Windows to follow.

Aside: in Dell’s marketing literature it warns consumers that a GB of hard drive space equals one billion bytes. That’s fine. Amusingly enough it makes exactly the same warning for installed memory, apparently not realizing that it is therefore inadvertently shipping customers 7.37% more memory than promised. Memory is the one time that a binary GB makes sense, but Dell’s lawyers apparently didn’t realize that.

Outlook, heal thyself

After upgrading I noticed that Outlook had decided that it needed to reindex all of my e-mail. I’m not sure why – I had been careful to ensure that Outlook was not running while the backup was being created. The next day I still couldn’t search my e-mails so I investigated, and I found this:

clip_image001[4]

Awesome. Outlook (or Windows search service, whatever, it’s Microsoft either way) decided that I was out of space. I never went below 30 GB free and I now had 363 GB free, but the indexing was stopped and it showed no inclination to restart itself.

Maybe this issue would have resolved itself after a reboot (have you tried turning it off and on again?) but it really shouldn’t be necessary. Effective e-mail search is practically a human right and it should be able to survive minor cataclysms like getting a bigger hard drive. Most users would not know how to correct this problem, and they shouldn’t have to. It should just work.

About brucedawson

I'm a programmer, working for Google, focusing on optimization and reliability. Nothing's more fun than making code run 10x faster. Unless it's eliminating large numbers of bugs. I also unicycle. And play (ice) hockey. And juggle.
This entry was posted in Computers and Internet and tagged , , , . Bookmark the permalink.

7 Responses to Hard Drive Upgrade

  1. jza says:

    Also the UNIX command line tools (df, free, etc.) still report in power-of-two units. Changing the default formats in them might break too many things, but they could start to offer a switch to show in power-of-ten units if the user desires.

    • brucedawson says:

      I took a quick look at df on my Ubuntu 12.04 system and it supports binary and decimal. The default is 1K-blocks (1,024 bytes) but you can specify –block-size=KB (1,000 bytes). The distinction between K/M/G/T and KB/MB/GB/TB is a bit subtle, but at least they give you the option. From the help:

      Display values are in units of the first available SIZE from –block-size,
      and the DF_BLOCK_SIZE, BLOCK_SIZE and BLOCKSIZE environment variables.
      Otherwise, units default to 1024 bytes (or 512 if POSIXLY_CORRECT is set).

      SIZE may be (or may be an integer optionally followed by) one of following:
      KB 1000, K 1024, MB 1000*1000, M 1024*1024, and so on for G, T, P, E, Z, Y.

  2. Adrian says:

    I once worked for a drive company. Drive sizes back then weren’t as simple as base 10 or base 2. For us 1 KB was 1024 bytes, and 1 MB was 1000 KB = 1,024,000 bytes. Ugh.

    The ISO metric prefix for kilo is k, not K. So I always assume KB is 1024 bytes and kB is 1000 bytes. Unfortunately, there’s no easy distinction for larger units unless you adopt the horrendous MiBi bytes nonsense.

    • brucedawson says:

      Hah — I’d never noticed that the metric prefix for kilo was lowercase ‘k’. Fascinating. I’ll have to try to remember to use that correctly.

      The 1.44 MB floppy disks of old are an awesome example of mixed units since that MB was defined as 1,024,000 bytes (http://en.wikipedia.org/wiki/Floppy_disk).

      I’d be happy to use MiB but there’s no sign that that is taking off.

  3. Pingback: Backin’ Up is Hard to Do | Random ASCII

  4. 8 hour backup, 5 hour restore? You look like you’re familiar with Linux, would clonezilla not have been simpler and quicker, especially as you could have both drives in the laptop at the same time? Clonezilla (via the ntfsclone utility) can go close to the speed of the drives. Even if that’s 50MB/s, that’s maybe an hour to do the clone.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s