Column values are used to identify individual entries in tables using the format:
columname="value", \ columnname="value", ...  | 
Where:
columname is the name of a table column
value is the contents of a particular cell within a column. That value is what identifies a table row. (When using column=value to create or modify table data, always enclose the value element in quotes.)
For example, suppose you had a hosts table that listed machine names and IP addresses:
Table 14-2 Example Hosts Table| 
 IP address  | 
 name  | 
 aliases  | 
|---|---|---|
| 
 129.146.168.4  | 
 altair  | |
| 
 129.146.168.119  | 
 deneb  | 
 
  | 
| 
 129.146.168.120  | 
 regulus  | 
 dnsmaster  | 
| 
 129.146.168.121  | 
 regulus  | 
 dnsmaster  | 
| 
 129.146.168.11  | 
 sirius  | 
In this example, your could identify the altair entry (row) in three different ways using the column=value of:
name=altair
address=129.146.168.4
name=altair,address=129.146.168.4.
But notice in the table above that the machine regulus is multi-homed and has two IP addresses. In that case, the column=value of host=regulus identifies two rows. To identify just the first regulus row, you would enter either:
address=129.146.168.120 or
address=129.146.168.120.,name=regulus,dnsmaster
Some nistbladm operations require that you enter a column=value pair for every column in the table.