OID Node Description

A node in the OID tree is either identified by the globally unique object identifier in dotted notation (e.g. 1.3.6.1.2.1.1.1) or by its globally unique fully qualified name (e.g. SNMPv2-MIB!sysDescr). 

Fully qualified names are composed out of the module name, which contains the definition of the MIB item, an exclamation mark and the label of the MIB item.

Simple names are also accepted in addition to fully qualified names. A simple name is a single label (e.g. sysDescr) or a sequence of labels separated by dots (system.sysDescr). Also, simple names are not necessarily unique and should be avoided in scripts where the set of known nodes in the OID tree is not known at implementation time.

It is possible to have numerical subidentifiers appended to a node name. These subidentifiers can specify an instance identifier, e.g. 

SNMPv2-MIB!sysDescr.0
or they can specify a path within the subtree defined by the node in the OID tree, e.g.
SNMPv2-MIB!system.1
or both, e.g.
SNMPv2-MIB!system.1.0
It is possible to have hexadecimal sub-identifier in an object identifier. A colon instead of a dot is used to indicate that the following sub-identifier should be read as a hexadecimal value.

For example, the object identifier 

1.3.6.1.4.1:627:74:75:62:73 
will be converted to
1.3.6.1.4.1.1575.116.117.98.115. 
You can also use a dot followed by a hexadecimal sub-identifier where the first two characters of the sub-identifier are 0x. It is also possible to mix both representation, e.g.
1.3.6.1.4.1.0x627:74:75:62:73.
It is suggested to use the object identifier notation as much as possible when referring to nodes in the OID tree. Globally unique names (like SNMPv2-MIB!sysDescr) are also a good and fast choice. 

Simple names may lead to ambiguities and can result is slower lookups if a name can not be found in the underlying hash table. Composed simple names (like system.sysDescr) make ambiguities less likely but they may cause substantial overhead