Linking with multiple back-ends (1 Viewer)

irish634

Registered User.
Local time
Today, 09:53
Joined
Sep 22, 2008
Messages
230
I'm not sure if this belongs here or in the "Theory and Practice..." area. Please move if needed.

My question encompasses a front end connected to multiple back ends. It is probably a mundane point, but reading a few articles didn't seem to touch on it.

By this, let me toss out that I am thinking of the personnel where I work when I say 95% have a hard time using email and MS Word, let alone mundane functions like find a file, etc. As such, I have to make my front-ends extremely user friendly.

With that in mind, when developers have multiple back ends, what are thoughts about the best way to relink when required? Even when there are 'more sophisticated' users? Security not an issue or is an issue (inasmuch as can be handled in access)?

Ex: Large corporation with a multitude of user ability. Large database with say 3 different back ends.

I was thinking about keeping a front-end table with each table name and UNC path, hard coding paths to corresponding table names, etc.

Personally I don't have a need for multiple back-ends, but the question came up while reading.
 

llkhoutx

Registered User.
Local time
Today, 08:53
Joined
Feb 26, 2001
Messages
4,018
Data is data is data, tables are tables are tables. regardless of where they reside. I modified a database which relinked back ends at every program start up. I don't see the necessity in doing that. One start up, check to see if a linked table will open without error, if no, no need to relink, if it dosen't open, relink,
 

irish634

Registered User.
Local time
Today, 09:53
Joined
Sep 22, 2008
Messages
230
Data is data is data, tables are tables are tables. regardless of where they reside. I modified a database which relinked back ends at every program start up. I don't see the necessity in doing that. One start up, check to see if a linked table will open without error, if no, no need to relink, if it dosen't open, relink,

Right, I agree. But regardless of where they reside and when they need to be re-linked, what is your preferred method of re-linking given the scenarios I listed (or others). For example, do you have your users 'browse' for each back-end?
 

essaytee

Need a good one-liner.
Local time
Today, 23:53
Joined
Oct 20, 2008
Messages
512
Right, I agree. But regardless of where they reside and when they need to be re-linked, what is your preferred method of re-linking given the scenarios I listed (or others). For example, do you have your users 'browse' for each back-end?

I have a default folder residing on the network which every man and his dog has access to. After front-end updates or when users wanting to change back-end databases the landing folder is the default folder. From there they navigate to their particular area.

In order to achieve this I have a preferences file in which one of the preferences relates to the 'default' data directory.

Steve.
 

gemma-the-husky

Super Moderator
Staff member
Local time
Today, 14:53
Joined
Sep 12, 2006
Messages
15,657
My databases have a local table which includes details of the linked tables that are required for the dbs to operate. If the tables are not correctly linked, the startup process relinks them in code.


do you mean your single database uses multiple backend databases, or do you mean you want to redirect a database to a different back end.

either way, the code is quite similar.
 

irish634

Registered User.
Local time
Today, 09:53
Joined
Sep 22, 2008
Messages
230
What most of my databases are used for, don't require multiple back-ends, but that could change. I was just reading some things about access and happened upon threads/articles talking about multiple back-ends.

It just got me thinking about how I would go about handling it given a scenario where I can't rely on someone to browse for the back-ends. The first thing that came to mind was exactly what you stated Dave and that was to house the information in a table to be called in the code.

To answer Dave's question, the scenario I am referring to in the discussion is having a single front end (for each user of course) but then having to link to several back-ends for whatever particular reason.

I try not to store files in a database but rather link them to a hidden folder. So my databases have really never grown to the point where I have found the need to separate the back ends into multiple files. The reading I came across just made me curious.
 

essaytee

Need a good one-liner.
Local time
Today, 23:53
Joined
Oct 20, 2008
Messages
512
I have a default folder residing on the network which every man and his dog has access to. After front-end updates or when users wanting to change back-end databases the landing folder is the default folder. From there they navigate to their particular area.

In order to achieve this I have a preferences file in which one of the preferences relates to the 'default' data directory.

Steve.

Forgot to mention that for one particular application in which users work on different back-ends, all my back-end data files I've purposely given them another extension (not .mdb) which is more meaningful. It also prevents the default opening of .mdb data files. When back-end files created by users, they only pick the name of the file, not the extension.
 

AOB

Registered User.
Local time
Today, 14:53
Joined
Sep 26, 2012
Messages
615
Interesting, I am just in the process of doing exactly this - incorporating the functionality to switch back-ends. I'm doing it for the purposes of archiving as my 'live' back-end will grow quite quickly but most data > 6 months old will not be relevant for 'day-to-day' processing (but will still need to be retained)

So I have written code to replicate the live back-end, copy qualifiying records over, then purge them from the live. The last piece in the puzzle is to provide the ability for users to switch to one of these archive files (on a read-only basis)

The method I am planning to use is to maintain a table of archives and allow the users to select one from a subform which will drive the relink. No need for any user to have visibility to the actual path of the archive, they can select by name (which by design will simply be a year)

But literally working on it right now so can't vouch for its suitability just yet!

By this, let me toss out that I am thinking of the personnel where I work when I say 95% have a hard time using email and MS Word, let alone mundane functions like find a file, etc. As such, I have to make my front-ends extremely user friendly

Personally, I always adopt a strategy of considering all my end users to be thick as two planks. That's not meant in a derogitory way; just if you can design something that even an idiot can use, you can be fairly sure the more intelligent ones will similarly have no such problems. It doesn't work the other way around. I think it's a good baseline from which to consider design!
 

Users who are viewing this thread

Top Bottom