Tips for creating/maintaining shared drive discipline?

Question

Shared drives seem to get so out of hand & it drives me crazy. Everyone organizes information differently, so a group directory where everyone has the right to create folders necessarily ends up with a lot of duplicate files, cryptic filenames, and assorted junk. Add to that constantly changing priorities and rapid turnarounds and you have a recipe for chaos. Other than locking down rights and having a librarian who keeps everyone in line, has anyone successfully transformed a messy directory where no one can find anything into an organized set of files that stays organized? And don't say buy a document management software package!

Answer 1

We had the same problem at my old job. Here's what we did (yes it did involve user cooperation, but when the boss was shown that we brought the backup time window back under the allowable downtime for both the file and email servers, didn't have to upgrade hardware, etc.,... he enforced it and that's the key - get top management to back you and you can solve the problem, don't and nobody will change.

Anyway....

1.) Departmentalized the file structure so each department had their own folder. Each department had full control of their department's folder but only read access to the other department's folders. Easy to do with group permissions and group membership.

2.) Each department could come up with whatever structure made sense to their function. The fact that each department has different needs, but still needs to communicate with other departments found its own balance - the structures naturally migrated to being quite similar all on their own.

3.) Instituted a policy where email was not for file transfers but to point to where in the file server to look for needed information. This cut down on the size of the email servers public and private stores in a huge way. We used exchange 5.5 where there are two huge files rather than lots of smaller files. This cut down on transfer errors and timeouts during the backup process.

4.) Instituted a dropbox setup. Each person had their own private dropbox in addition to a shared public dropbox. If you needed to get a file to someone, you just dragged it into that person's dropbox and sent an email telling them to look. The files were purged when they were 7 days old based on individual creation dates.

The drop boxes were read only for the public, write only for those putting a file in there so there were no accidental deletions.

For private drop boxes, they were write only for everyone but the owner who had full control.

5.) naming conventions - we went with name of file - dept abbreviation+title+version. Worked well enough once people understood how to search for files and actually stuck with the convention.

6.) keep the old files structure in place, but read-only so nothing is lost, but a migration to the new system is enforced.

7.) understand that there will be no immediate payoff - it takes time to purge the effects of the chaos still in the system. Management and the users have to understand this or they won't buy into the new system.

This system worked very well, but again, it wasn't perfect. There were still occasional issues, but they were far smaller than the problems we had before.

Again, I can't stress enough the importance of upper management being behind you on this - if they aren't, you will get squashed.

This is really more of a political issue than a technological one. People don't like change so to convince them TO change, there has to be a reason that benefits THEM personally rather than just a technological justification. Remember the technology is there for the people, not the other way around - hard to do when you are the one that stays late at night fixing everyone else's easily preventable problems knowing they went home early because of the same problem.

Try to put it in terms of productivity and money - anything that lowers productivity, lowers cost, raises productivity or raised costs are what gets upper managements attention - put it in terms that are important from their perspective and they will either back you or accept the problem as is. If they give you the go ahead, they have agreed to back you up. If they don't, they have agreed its not your fault.

When we did this change, backup window went from 13 hours to 5 hours for the entire network. Duplicate files just about disappeared (saved money in avoiding upgrades) File find time went down in a huge way. Inaccurate information issues disappeared (a single source file change was instantly aware to all the users rather than one person changing their copy while everyone else was using their own personal copy. No way to measure that, but when its a change in say return policy for product, that can have huge effects - particularly when customers DO talk to one another and get different policies because the staff isn't on the same page.