Modeling issue

Hi all,
I'm trying to model this situation but I have no idea how to handle it:

Hierachy: NODE - OBJS

Dimension Table for NODE (100 nodes)
Dimension Table for OBJS (1000 OBJS for each NODE) --> 100*1000 rows

Code:


NODEID  NODENAME PAR1 PAR2
1      NODO1    aa  bb
2      NODO2    cc  dd
2      NODO3    ee  ff


Code:


OBJS_ID  OBJNAME NODEID  PARA PARB
1        OBJA    1      aa    bb
2        OBJX    1      a1    b1
3        OBJW    1      a3    b2
4        OBJB    1      aa    bb
5        OBJA    2      aa    bb
6        OBJB    2      aa    bb
7        OBJY    2      a2    b2
8        OBJF    2      a11  b2
9        OBJA    3      aa    bb
10      OBJB    3      aa    bb
11      OBJZ    3      a3    b3
12      OBJS    3      a3    b22


I also need to keep track of history of NODE and OBJS.

I know that there are lots of OBJS that are common (they have same name and attributes) for all Nodes
In the example OBJS_ID (1,5,9) refer to (NODE1, NODE2, NODE3) with the same OBJ_NAME and params. Considering that there is a great percentage of OBJ_ID referring to all NODE_ID with all the same params what is the way to model this situation?

Thanks
Riccardo

ric79 wrote:I know that there are lots of OBJS that are common (they have same name and attributes) for all Nodes
In the example OBJS_ID (1,5,9) refer to (NODE1, NODE2, NODE3) with the same OBJ_NAME and params. Considering that there is a great percentage of OBJ_ID referring to all NODE_ID with all the same params what is the way to model this situation?

Forget about percentages when designing a data model. If it happens once it may as well happen all the time if the model is supposed to support it.

What is the nature of an object? Is OBJA the same object regardess of which node it is associated with? Or, are they different objects and are uniquely identified in context with the node? If OBJA is the same object, are the parameters independent of the object (values can change from node to node) or are they attributes of the object (i.e. always the same for an object regardless of the node, if the parameters of an object change, they change for all associated nodes)?

Thanks for reply.

I.E:
NODE is a TLC node (MSC, TRA, GTW, SGT, SGU, ecc.)
OBJS are Measures Types (OBJ1:CPU measures, OBJ2:RouteOutgoing, OBJ3:ErrorCodes)

(TYPE1) OBJ1 PK:CPU Name It Has 2 attributes in the FACTTable: CPUCharge, CPUErrors
(TYPE2) OBJ2 PK: RouteOutgoing It Has 3 attributes in the FACTTable: destinationErrors, destinationPackets, destinationSetupTime
(TYPE3) OBJ3 PK: ErrorCause It has 1 attribute in the FACTTable: numErrorWithThisCause

OBJ2 lists all the rout of a Node, so it's clear that a Rout can belong just to a Node.
The ErrorCause are all the same for all the nodes. (ErrorCause=1 means "no memory" both for NODE1 and NODE2, ErrorCause=2 means "segmentation fault" for all nodes, etc..)
There are 6000 ErrorCauses X 100 Nodes.
Do I need to anagraph 600000 different ErrorCauses or is there a way to anagraph just the 6000 OBJ3 and spreading in such a way this info on all nodes:

Code:

NODE1  ---  CPU Name      --- CPU1
                            --- CPU2
        ---  RouteOutgoing  --- RoutA1   
                            --- Rout2S   
                            --- RoutEd     
        ---  ErrorCause    --- Cause1
                            --- Cause2
                            ....
                            --- Cause6000
...
NODE100 ---  CPU Name      ---  CPUC
                            ---  CPUdee
        ---  RouteOutgoing  --- RoutA12ww3
                            --- Rout2sw
        ---  ErrorCause    --- Cause1
                            --- Cause2
                            ....
                            --- Cause6000

Riccardo