Planning and Implementing VLANs with HP-UX
Table Of Contents
- Planning and Implementing VLANs with HP-UX
- Table of Contents
- About This Document
- What is VLAN?
- VLANs on HP-UX
- Features and Advantages
- Implementing VLANs on HP-UX
- Priority and Class of Service (CoS)
- IP ToS and 802.1p Conversion—End-to-End Class of Service
- Typical Customer Configurations
- Using HP-UX VLANs with HP Auto Port Aggregation (APA)
- Using HP-UX VLANs with HP Virtual Machines (HPVM)
- Future HP-UX VLAN Feature Additions
that hub must belong to the same VLAN or set of VLANs, thereby restricting the benefits
of VLANs.
• A common misconception is that because multiple IP subnets can share a single switched
infrastructure using VLANs, switching can replace routing in the network. Remember that
VLAN is strictly a Data Link Layer (Layer 2) technology. You must use routers for
communication between IP subnets, even in a VLAN. [In the case where you are using a
VLAN-aware Layer 3 (routing) switch (mentioned previously as one of the available types
of VLAN-aware switches), you do not need a separate router, because VLAN-aware routing
switches incorporate both the Layer 2 and Layer 3 functions.]
VLAN Tagging
As mentioned previously, you can implement VLAN functionality via explicit frame tagging by
switches and end stations. Network switches and end stations that know about VLANs are said
to be VLAN aware. Network switches and end stations that can interpret VLAN tags are said to
be VLAN tag aware. VLAN-tag-aware switches and end stations add VLAN tags to standard
Ethernet frames–a process called explicit tagging. In explicit tagging, the end station or switch
determines the VLAN membership of a frame and inserts a VLAN tag in the frame header (see
Figure 2), so that downstream link partners can examine just the tag to determine the VLAN
membership.
Tagging has several advantages—VLAN association needs to be applied only once at an end
station or at an edge switch, so that downstream switches all the way to the destination are
relieved of the burden of classifying frames. Tagging at end stations is particularly beneficial
because the overhead of frame classification is distributed.
Standards and Interoperability
IEEE 802.1Q specifies the architecture for VLAN tagging—tag format, tag insertion, and tag
stripping. The IEEE 802.1Q tag (shown in Figure 2 for an Ethernet frame) also has a provision
for priority encoding. The 3-bit priority field in the tagged frame carries priority information.
IEEE 802.1p (later incorporated in IEEE 802.1D) has standardized this priority encoding.
VLAN Trunking
Switches that implement only port-based VLAN can support only one VLAN per port. However,
if they are tag aware (also called Q-compliant), they can support multiple VLANs per port— one
untagged VLAN and multiple tagged VLANs. If a frame doesn’t have an explicit VLAN tag, it is
automatically assigned the “untagged VLAN ID” or the “default VLAN ID.” An inbound frame
that is tagged has its VLAN ID in the frame header. Some switch vendors refer to the ability of
handling multiple tagged frames per port as VLAN trunking.
If the end node is a VLAN-aware server, the LAN card port on the server can support multiple
VLANs per port, as shown in Figure 3.
10