1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
|
int: N_CLOUD_REGIONS; % number of cloud regions
int: N_ATTRIBUTES; % number of capability related attributes
array[1..N_ATTRIBUTES] of float: W_ATTRIBUTES; % weights of each attribute
int: N_METRICS; % number of dynamic capacity metrics of interest
array[1..N_METRICS] of float: W_METRICS; % weights of each capacity metric
int: CUST_TYPE; % customer type, 0 = regular, 1 = silver, 2 = gold
enum CUST_TYPES = { STANDARD, SILVER, GOLD };
enum ATTRIBUTES = { CORE_DC, DIRECT_CONN, MIN_GUARANTEE, SRIOV };
enum METRICS = { AVG_UTILIZATION, PEAK_UTILIZATION };
% whether a cloud region has the corresponding capability -- data will be customer specific
array[1..N_CLOUD_REGIONS, 1..N_ATTRIBUTES] of int: capabilities;
array[1..N_CLOUD_REGIONS, 1..N_METRICS] of float: utilization; % how much capacity is already utilized (fraction)
var int: s; % target cloud region (solution to the problem)
% custom constraints
constraint capabilities[s, CORE_DC] = 1; % hard constraint: has to be placed in CORE DC
constraint utilization[s, AVG_UTILIZATION] <= 0.85; % hard constraint: need some capacity available
% custom soft constraint for gold customers -- give a large weight to direct connection
var float: additional_obj = bool2int(CUST_TYPE = GOLD) * capabilities[s, DIRECT_CONN] * 1000;
% TODO: global constraints (such as data validation)
% Objective for utilization
var float: obj_utilization = sum(k in 1..N_METRICS) ( W_METRICS[k] * (1 - utilization[s, k]) );
% Objective for capabilities
var float: obj_capabilities = sum(k in 1..N_ATTRIBUTES) ( W_ATTRIBUTES[k] * capabilities[s, k] );
% Overall objective function
var float: obj = obj_utilization + obj_capabilities + additional_obj; % can later add weights to each...
solve maximize obj;
output ["Cloud Region: ", show(s), "\n", "Objective function value: ", show(obj), "\n", "Customer type: ", show(CUST_TYPE), "\n"];
|
