OK let’s say I have a 2tb hard drive
And I create a bunch of partitons
OK my a drive is reserved for floppies (is that still true) as is b drive
OK my main drive is C and the CD is D drive
Now I create on my hard drive
22 more partitons and assign the letter E through Z to each of the partitions.
Now I go to create a 23rd partiton, what happens?
I never thought about this before, and of course probably no real reason to do it in reality, but I was just wondering.
It won’t be mapped, in my experience with XP. At my old job, we had certain network drives that were mapped as E: and F:. If I was using a machine that had lots of local drives, some of which could also be mapped as E: and F:, either the local drive or the network drive would show up for each drive letter, depending on what things got mapped first by the startup scripts that accompanied my login. I could still go in to Disk Manager (or whatever it was called), see all available partitions, and reassign drive letters as need be.
I don’t know what the situation is in Vista or Win7.
I’m guessing you mean on Windows. Simple answer, before Windows 2000, you were pretty much screwed. For Windows 2000 and beyond, with NTFS, you can create a ‘junction point’, which can map an entire drive to a an empty directory of another.
You can also refer to it directly, and map it, but that takes a bit more work. Gah… I was looking for some resources on this, but I am at the wrong computer. A good place to start is:
Edit per post just before mine -
With Windows, you do not need to actually map network drives. They are fully accessible using the share name, without mapping. The only reason to map them to a physical A:-Z: ‘Disk’ would be for old, or poorly behaved applications.
Any true 32 bit application (again, unless poorly behaving) should understand the “\MyServer\TheShare\TheFile” designation, but the OP was asking about a single, local drive partitioned off in a manner that would really not make much sense today.
Edit whilst editing - You can also use something like VMWare to create whatever you like, with the restriction that you have to run a virtual machine for each set of ‘physical’ mappings.
Well, you also need to map them to log on to a folder as a different user without logging out.
For instance, I usually go to network folders just by typing the path. But in some cases, I need to log in with account A* to access a folder that is only accessible from account Z. You need to then map a network drive to the folder account Z has access to and use the account Z credentials.
*Account A is an administrator account that has no access to account Z’s folders.
Win7 will let you assign A: and B: to harddrives, including the boot drive. VERY BAD IDEA though, as doing so opens to you all kinds of compatibility issues, since many programs hardcoded a: and b: to = floppy drive and thus will not install to/run from those letters.
In the case of Win7, if you have more than 26 devices (those stupid all in one card readers can make it add up really fast too), you need to go into the Computer Management control panel and reassign the letters, since devices will just stop showing up after A-Z are claimed.
Thanks for the info, interesting stuff.
Markxxx, may I ask why you ask that question?
Just that it’s generally way more efficient to keep big drives (even partitions, though there is little reason to partition now).
I mean, why would you WANT to partition a 2tb drive into 10gb sections?
For the most part, I work on agglomerating my drives into a single virtual huge one (Databases, ya know). I keep (even at home) a relatively small drive for booting, and the rest is free-for-all
Regards
I was just wondering, I had no intention of doing it. I have a 500mb “C” drive and I have a 2tb drive I use for storage. 1tb for music “M” Drive and 1tb for TV (“T” drive).
I was just defragging the drive. Took long while to defrag the 1tb music drive since it’s never been done and it’s 6 month.
Anyway I got to wondering about drives in general.
I tried to Google the question but I couldn’t ask it in Google so I got a decent answer.
As a data point, two jobs back I encountered the strange situation where a usb drive was automatically assigned to d:\ drive, even if d was already taken. You could go into disk manager and re-assign it, which turned out to be the workaround I used, however it struck me as very very odd that Windows (XP, in case it’s significant) would recognize it then assign it to an already in use drive letter. I never did figure out if it was a problem with the configuration (certainly possible) or a windows bug.
Interesting question, by the way. It never occurred to me before, though it’s a perfectly reasonable question. I work in QA now, I may have to use that for some test scenarios.
You mean PCs still have that “mapped letters for drive” thing going?
I guess I knew that they still could but do they have to? I thought that their existence nowadays was just for legacy and backwards compatibility purposes for some programs or something.
Are you seriously limited to no more than 26 local volumes?
By default, yes. But, as mentioned, you can manually mount the drive to an empty folder.
What a hindrance! Why would this surprise you? If you really need more than 26 volumes, why the hell are you using Windows (or a Mac, even) anyway?
My Commodore 64 wouldn’t handle it?
OK maybe I’m missing the obvious here, but what’s the point of defragging a single partition on a hard drive?
I realise that if both of the partitions in your example are relatively empty, then defragging one will collect all of it’s data together. If on the other hand both partitions are relatively full, then defragging one at a time wont helpmuch because fragmentation from the other drive will reduce the effectiveness.
Wouldn’t it be more effective in the long run to have your music and videos as two folders on one drive? Each run of defragging would have to handle more data, but it would defrag more thoroughly.
Your typical defragger just does the whole drive, it doesn’t do specific folders or anything. By partitioning out the drive, you could limit it so that it doesn’t have to run for 26 hours to complete.
This is a long standing bug. USB devices will take the next available non-physical drive letter, but if its already mapped as a network drive, they’ll ignore the mapping and just fail to work without even producing an error. I’ve only seen this on XP. Not sure if Vista/7 has this problem.
You can map to folders or use iSCSI or WebDav. You can have thousands of mounts.
Yes and no. If you make a partition just for apps and OS, which doesn’t get changed a lot, then you’re not going to have many fragmentation issues. When it comes to folders that get updated constantly like music, temp, etc then you’re going to get all these fragmentation issues and will need to defrag the drive.
Some people make a system partition that doesnt get fragmented as fast and a data partition which does get fragmented quickly. This limits fragmentation on the OS, which is using all sorts of temp files, caches, etc and is suseptible to performance issues if youre also writing data files and filling up all these blocks and encouraging fragmentation.
Not to mention, that first partition will always be on the first part of the disk, which is always faster than the rear part.
No, as mentioned in post #3, they are not really necessary except for occasional backwards compatibility with stubborn applications (well OK, you probably do need a drive letter for your boot/system disk-- I bet Windows wouldn’t like it if it couldn’t find C: ). These days the UNC format is preferred (\hostname\sharename\path), or, as mentioned, you can use moint points like in Unix. It has been this way for a good 10 years.
Mind you, people seem very fond of drive letters. I try to explain to my users that they don’t need to map F: or whatever to their favourite network share, but some of them seem more comfortable doing it the way they’ve always done it.
It’s still used a lot in the corporate world. Personal share is your P: drive, and other letters are for your group shares. Users are more comfortable knowing they have to open their K:/ drive to get to the TPS report, than having to navigate to \server\groupshare\dumbreports to get to the TPS folders. It also makes management of the volume space easier for administrators if the users don’t have to know “where” the data is. Doubly so if you need to change the server name for some reason.
Thanks for that. For some reason it never occured to me that partitioning a drive isolated a contiguous block of that drive.