Server Database

[Bill]

Hi All,

I am looking to redesign our server database to deal with clusters and
virtual servers better than the existing one does.

My problem is that whatever design I try I am constantly drawn back to the
'Physical Sever' table with say the asset ref (or serial number) as tie
primary field and then another field in 'Physical Server' to identify what
cluster the hardware is a member of.

I then want to use cluster name as the primary field in another table that
identifies what service the cluster hosts. The trouble is that the only way
that I can get the data as primary is to use a unique or group query . The
problem with that though is that the relationship does not allow cascade
update with any secondary tables which is such a powerful facility.

Anyone any ideas how I go about this?

Regards.
Bill.

 

[Jeff Boyce]

Bill

You've described something of a "how", how you are trying to do something.

I don't have a clear enough view of "why" yet to offer specific suggestions.
Since it all starts with the data, could you post back a description of the
data/domain you are working in?

--
More info, please ...

Jeff Boyce
Microsoft Office/Access MVP


Microsoft IT Academy Program Mentor
http://microsoftitacademy.com/

Microsoft Registered Partner
https://partner.microsoft.com/

 

[Bill]

Jeff,

Sorry, what do you mean by data/domain?

Regards.
Bill.

 

[Bill]

Ok, guessing at what you are asking.

In the table for the physical servers I have the a unique reference for the
item of hardware. That will record in another field called 'cluster name'
the name of the server cluster that it is a member of. If the item of
hardware is not in a cluster in will simply have its own name in the
'cluster membership' field.

In another table I want to record the services that are delivered by either
the single server or by the server cluster. This table needs to have the
name of the server cluster as a unique primary field. The problem is that in
the 'physical' table the cluster name will appear many times. I don't want
to enter the names separately and would like the to be related with a
cascade update etc available.

Hope this helps.

Regards.
Bill.

 

[Dennis]

First, what's a "primary field?" Second, your issue seems easy enough to deal
with. How are you entering data? Have you designed a form/subform? What are
you talking about in terms of "cascade updating?" It seems that you only have
two tables; one for actual equipment and one for the "server" (cluster) that
each piece of equipment is assigned to. (Actaully, now that I think about it,
you'd have a third table where you track the "service" that each
server/cluster is assigned to.)

So what's the problem? This isn't a complex design at all if I understand
you correctly.

 

[Jeff Boyce]

See comments in-line below...

Ok, guessing at what you are asking.

In the table for the physical servers I have the a unique reference for the
item of hardware. That will record in another field called 'cluster name'
the name of the server cluster that it is a member of. If the item of
hardware is not in a cluster in will simply have its own name in the
'cluster membership' field.

It sounds like you have a table with:

RowID
ServerName
(?)SerialNumber
(?)Manufacturer
MemberOfClusterName

Why bother putting ANYTHING into [MemberOfClusterName] when the physical
server is not ...? That is, what are you doing this to help you accomplish?

In another table I want to record the services that are delivered by either
the single server or by the server cluster.

So, you really don't care about physical or "cluster" servers specifically,
only that these represent "Service Providers"?

This table needs to have the
name of the server cluster as a unique primary field.

Why? Why the constraint? A unique primary key field does NOT have to be a
value that a human can recognize? What business need are you solving by
including this as a requirement?

The problem is that in
the 'physical' table the cluster name will appear many times. I don't want
to enter the names separately and would like the to be related with a
cascade update etc available.

Access is a relational database. How 'bout if you set up a ClusterNames
table and use the ID from that table to show which cluster a server belongs
to? Instead of [MemberOfClusterName], you'd use [ClusterID].

Hope this helps.

Regards.
Bill.

Good luck!

--
Regards

Jeff Boyce
www.InformationFutures.net

Microsoft Office/Access MVP


Microsoft IT Academy Program Mentor
http://microsoftitacademy.com/