Random order of one pandas.DataFrame with respect of other

I have the following structure:

data_Cnx = pd.read_csv(path_Connection,sep='\t',header=None)
data_Cnx.columns = ["ConnectionID"]
data_Srv = pd.read_csv(path_Service,sep='\t',header=None)
data_Srv.columns = ["ServiceID"]

that can be visualized as the following:

print(data_Cnx)
      ConnectionID
    0   CN0120
    1   CN0121
    2   CN0122
    3   CN0123
    4   CN0124
    ...           ...
    20   CN0166
    21   CN0167
    22   CN0168
    23   CN0171
    24   CN0172
    [25 rows x 1 columns]      
  
print(data_Srv)
       ServiceID
    0   ST030
    1   ST030
    2   ST030
    3   ST030
    4   ST030
    ...          ...
    20  ST040
    21  ST040
    22  ST040
    23  ST050
    24  ST050
    [25 rows x 1 columns]

Literally, each element from data_Cnx corresponds to a parallel element in data_Srv, respecting the order. For instance:

CN0120 corresponds to ST030
CN0121 corresponds to ST030
....
CN0166 corresponds to ST040
CN0167 corresponds to ST040
...
CN0171 corresponds to ST050
...

I would like to have another structure or different data_Cnx and data_Srv in which the order of data_Cnx can be randomized, but always in respect of what corresponds in data_Srv. For instance:

The data_Cnx and data_Srv can be visualized as the following:

print(data_Cnx)
      ConnectionID
    0    CN0120
    1    CN0168
    2    CN0156
    3    CN0133
    4    CN0161
    ...           ...
    20   CN0121
    21   CN0143
    22   CN0127
    23   CN0151
    24   CN0132
print(data_Srv)    
    [25 rows x 1 columns]        ServiceID
    0   ST030
    1   ST040
    2   ST070
    3   ST010
    4   ST040
    ...          ...
    20  ST030
    21  ST050
    22  ST030
    23  ST070
    24  ST010
    

I was thinking of using randn, but obviously it uses integers as parameters. Do you have an easier idea of how can this be implemented?