skip to main | skip to sidebar
Showing posts with label CCNP Switching. Show all posts
Showing posts with label CCNP Switching. Show all posts

Wednesday, November 4, 2009

CCNP BCMSN (642-812) Online Lab

0 comments

CISCO Catalyst 6500 Labs

1 comments

Friday, July 18, 2008

8-6. Router Redundancy with HSRP

0 comments
· Route processors in the same or another chassis can share redundant gateway addresses on a VLAN by using the Hot Standby Router Protocol (HSRP).
· Route processors sharing a common HSRP IP address must belong to the same HSRP group number.
· The HSRP address appears on the network with a special virtual MAC address00-00-0C-07-AC-XX, where XX is the HSRP group number (0 to 255). The hosts on the HSRP VLAN use this MAC address as the default gateway.
· Although HSRP is enabled on an interface, each route processor still maintains its own unique IP and MAC addresses on the VLAN interface. These addresses are used by other routers for routing protocol traffic.
· When an HSRP group is enabled, the highest-priority HSRP device at that time becomes the active router, whereas the second-highest-priority stays in the standby state. All other HSRP devices in the group maintain a "listening" state, waiting for the active device to fail. A new active router election occurs only when the active device fails. The previous active router (having the highest priority) may reclaim its active role by preempting the other HSRP routers in the group.
· HSRP devices communicate by sending a hello message over UDP at multicast address 224.0.0.2. These messages are sent every 3 seconds by default.
· Devices on a VLAN use the HSRP address as their default gateway. If one of the HSRP devices fails, there will always be another one to take its place as the default gateway address.

8-5. MSFC Redundancy with Configuration Synchronization

0 comments
· In config-sync mode, both MSFCs are active at all times; all interfaces and routing processes are available and active on both modules.
· One MSFC is "designated" and maintains the master copies of startup and running configurations. The other is "nondesignated" and receives its configurations from the designated module.
· The nondesignated MSFC can be monitored through an EXEC session, but doesn't allow the configuration mode.
· Config-sync mode allows an immediate failover because both MSFCs are always active. All VLAN interfaces are up, and any configured routing protocols are active on both modules.
· For MLS, either of two redundant MSFC modules can be involved in setting up a flow cache entry. The PFC is aware of both RP modules.
· For CEF, only the designated MSFC downloads the FIB and adjacency table information to the other modules. The designated MSFC is the one that initializes first, in the lowest module slot position.
· HSRP should be used to provide a redundant gateway address on each VLAN. Both MSFCs share a common gateway address, while operating their own unique interface addresses. (See section "8-6: Router Redundancy with HSRP" for more information about HSRP.)
· After config-sync mode has been enabled and is active, any configuration changes that are made on the designated MSFC are automatically synchronized with the nondesignated module:
- Whenever you enter the write mem or copy source startup-config commands, the startup configuration is updated across MSFCs.
- Whenever you enter the copy source running-config command, the running configuration is updated across MSFCs.
- As you enter commands in configuration mode, they are also sent to and executed on the nondesignated MSFC.

8-4. MSFC Redundancy with Single Router Mode

0 comments
Only the Catalyst 6000 supports redundant supervisor engines with redundant route processors (MSFC/MSFC2).
Only one MSFC is designated and active; the other is booted up, synchronizes its configuration with the active MSFC, starts any routing protocol processes, but keeps all interfaces in the "line down" state so that no network traffic is exchanged.
Only the designated and active MSFC is involved in creating MLS flow cache entries for the switching engine and in creating and downloading the CEF FIB and adjacency tables to the PFC2 switching engine.
In the event of a failure, single router mode (SRM) causes the nondesignated MSFC to bring its interfaces up and allow its routing protocols to converge. During this time, the existing Layer 3 switching information supplied by the failed MSFC continues to be used by the designated Supervisor PFC until the new MSFC can provide an update.
Both MSFC modules must run the same Cisco IOS Software imageSoftware Release 12.1(8a)E2 or later.
The Supervisor modules must run in high-availability mode to support SRM, using Supervisor Engine Software Release 6.3(1) or later.
Hot Standby Router Protocol (HSRP) is not necessary in SRM, because only one MSFC is active in the chassis at any one time. However, you should use HSRP if you have other MSFC or Layer 3 devices that have a presence on the same VLANs as the SRM MSFC. These other devices can then act as redundant gateways on those VLANs.

8-3. NetFlow Data Export

0 comments
Traffic statistics from Layer 3 switching can be gathered and sent to an external application for collection and analysis. This is done through the NetFlow Data Export (NDE) facility.
Switches using MLS for Layer 3 switching can send data about expired flows using NDE. This is a natural extension of MLS because the switch uses flow cache data.
Switches using CEF do not inherently use a flow cache, and therefore can't offer statistics through NDE. The Catalyst 6000 PFC2/MSFC2, however, keeps a NetFlow cache independent of the CEF process, strictly for exporting flow data with NDE.
NetFlow data can be sent as several versions:
- NDE version 1 Used in legacy systems; data record includes specific information about the IP traffic flow and the interfaces used to forward it.
- NDE version 5 Adds a sequence number to prevent lost UDP datagrams, and the Border Gateway Protocol (BGP) autonomous system (AS) number for the flow.
- NDE version 7 Used to report flow data from Catalyst switches. Version 7 is not supported on a Catalyst 6000 MSFC.
- NDE version 8 Used to report aggregate flow data from routers, Catalyst 5000 with NFFC, and Catalyst 6000 running MLS or CEF. Version 8 is not supported on a Catalyst 6000 MSFC.
NDE will export flow statistics according to the MLS flow mask that is used by the switch. To see detailed flow records, use a "full" flow mask.

8-1. Multilayer Switching

26 comments
Multilayer switching (MLS) performs Layer 3 switching by combining separate routing and switching functions on different switch modules.
MLS is supported on the Catalyst 6000 by the MSFC (route processor) and the PFC (Layer 3 switching engine). On the Catalyst 4000 and 5000 platforms, the route processor can be an external router.
MLS can perform Layer 3 switching for IP, IP multicast (also known as IP Multicast MLS or MMLS), and IPX traffic.
MLS operation consists of these steps:
- The route processor (RP) routs the first packet in a traffic flow.
- The switching engine (SE) sets up an MLS cache entry for the flow based on the first packet (a "candidate" packet).
- When the SE sees the return packet from the RP, the MLS cache entry is completed with source and destination information. For the duration of the traffic flow, subsequent packets are switched at the SE.
- When the SE switches flow packets, the SE also rewrites the source and destination MAC addresses, the IP time-to-live (TTL), and both Layer 2 and Layer 3 checksum values. This is done in hardware, as if a traditional router had forwarded the packets.
- The MLS cache entry for a flow is deleted when the connection is closed or after an aging timer expires.
MLS builds its flow cache based on the following:
- IP Destination address, source and destination addresses, or source and destination address and port numbers ("full flow")
- IP multicast Source address, source VLAN number, and destination multicast group
- IPX Destination address
MLS can report on traffic flow statistics through the use of NDE. Refer to section "8-3: NetFlow Data Export" for more information.
-2. Cisco Express Forwarding
Cisco Express Forwarding (CEF) handles all packet forwarding in hardware, for all packets in a flow.
CEF is implemented on the Catalyst 2948G-L3, 4908G-L3, 4000 Supervisor III, and 3550 series switches. It is also implemented on the Catalyst 6000 as a cooperation between the PFC2 Layer 3 switching engine and the MSFC2 route processor module.
A route processor runs routing protocols and populates the following tables:
- The normal routing table A table of routes and next-hop destinations as determined by the routing protocols, administrative distances, metrics, and so on.
- The Forwarding Information Base (FIB) Every known route is represented in the FIB as a hierarchical tree structure. Longest-match routes can then be quickly looked up in hardware, pointing to the next-hop entry in the adjacency table.
- The adjacency table Every next-hop router address and Address Resolution Protocol (ARP) reply that is discovered is entered into the adjacency table, giving an efficient Layer 3-to-Layer 2 forwarding lookup.
CEF supports high-performance switching of IP, IP multicast, and IPX traffic.
CEF can switch packets over up to six equal-cost paths to a common destination.
CEF can use Reverse Path Forwarding (RPF) to make sure packets arrive on interfaces that are the best return paths to the source. This can be used to detect forged or spoofed addresses in received packets, in the case of some malicious activity.
IP multicast traffic is switched by CEF only for multicast groups within 225.0.0.* through 239.0.0.* and 224.128.0.* through 239.128.0.*. CEF will not switch anything in 224.0.0.* because those addresses are reserved for routing protocols and must be flooded to all ports that are forwarding in a VLAN.
When the route processor creates the FIB, the FIB information is downloaded and used by the switching engine hardware. On a Catalyst 6000, the FIB is downloaded from the MSFC2 to the PFC2 module, as well as any distributed forwarding cards (DFCs) that are present.
In addition to the CEF tables, a NetFlow forwarding table (identical to that of MLS) is independently generated just to provide flow-based accounting information. This information can be exported to external applications. See section "8-3: NetFlow Data Export" for more information

7-3. STP Convergence Tuning

0 comments
· STP bases its operation on several timers. Usually, the default timer values are used for proper STP behavior. The defaults are based on a network diameter of seven switches but can be adjusted for faster convergence times.
- The Hello Timer triggers periodic hello messages to neighboring switches.
- The Forward Delay timer specifies the time a port stays in each of the listening and learning states.
- The MaxAge timer specifies the lifetime of a stored BPDU received on a designated port. After the timer expires, other ports can become designated ports.
· BPDUs are expected at regular intervals. If they are delayed beyond the lapse of an STP timer, topology changes can be triggered in error. This condition can be detected with the BPDU skewing feature.
· STP PortFast allows ports that connect to hosts or nonbridging network devices to enter the forwarding mode immediately when the link is established. This bypasses the normal STP port states for faster startup, but allows the potential for bridging loops to form.
· STP UplinkFast is used only on leaf-node switches (the ends of the ST branches), usually located in the access layer. The switch keeps track of all potential paths to the root, which are in the blocking state.
- When the root port fails, an alternate port is brought into the forwarding state without the normal STP port state progression and delays.
- When UplinkFast is enabled, the bridge priority is raised to 49152, making it unlikely to become the root bridge. All switch ports have their port costs increased by 3000 so that they won't be chosen as root ports.
- When an alternate root port comes up, the switch updates upstream switches with the new location of downstream devices. Dummy multicasts are sent to destination 01-00-0C-CD-CD-CD that contains the MAC addresses of stations in the bridging table.
· STP BackboneFast causes switches in the network core to actively look for alternate paths to the root bridge in case of an indirect failure.
- When used, this feature should be enabled on all switches in the network. Switches use a request-and-reply mechanism to determine root path stability, so all switches must be able to participate.
- BackboneFast can only reduce the convergence delay from the default 50 seconds (20 seconds for the MaxAge timer to expire, and 15 seconds in both listening and learning states) to 30 seconds.

7-1. STP Operation

0 comments
· STP detects and prevents Layer 2 bridging loops from forming. Parallel paths can exist, but only one is allowed to forward frames.
· STP is based on the IEEE 802.1D bridge protocol standard.
· Switches run one instance of STP per VLAN with Per-VLAN Spanning Tree (PVST). PVST between switches requires the use of Inter-Switch Link (ISL) trunking.
· For IEEE 802.1Q trunks, only a single instance of STP is allowed for all VLANs. The Common Spanning Tree (CST) is communicated over VLAN 1.
· PVST+ is a Cisco proprietary extension that allows switches to interoperate between CST and PVST. PVST bridge protocol data units (BPDUs) are tunneled over an 802.1Q trunk. Catalyst switches run PVST+ by default.
· Multiple Instance Spanning Tree Protocol (MISTP) is also a Cisco proprietary protocol that allows one instance of STP for one or more VLANs via a mapping function. This allows faster convergence with a lower CPU overhead and fewer BPDUs. MISTP discards PVST+ BPDUs.
· MISTP-PVST+ is a hybrid STP mode used to transition between PVST+ and MISTP in a network. BPDUs from both modes are understood and not discarded.
· Multiple Spanning Tree (MST), based on the IEEE 802.1s standard, extends the 802.1w Rapid Spanning Tree Protocol (RSTP) to have multiple STP instances.
- MST is backward-compatible with 802.1D, 802.1w, and PVST+ STP modes.
- Switches configured with common VLAN and STP instance assignments form a single MST region.
- MST can generate PVST+ BPDUs for interoperability.
- MST supports up to 16 instances of STP.
· Switches send BPDUs out all ports every Hello Time interval (default 2 seconds).
· BPDUs are not forwarded by a switch; they are used only for further calculation and BPDU generation.
· Switches send two types of BPDUs:
- Configuration BPDU
- Topology change notification (TCN) BPDU
NOTE
BPDUs are sent to the well-known STP multicast address 01-80-c2-00-00-00, using each switch port's unique MAC address as a source address.

STP Process
1. Root bridge election The switch with the lowest bridge ID becomes the root of the spanning tree. A bridge ID (BID) is made up of a 2-byte priority and a 6-byte MAC address. The priority can range from 0 to 65535 and defaults to 32768.
2. Root port election Each nonroot switch elects a root port, or the port "closest" to the root bridge, by determining the port with the lowest root path cost. This cost is carried along in the BPDU. Each nonroot switch along the path adds its local port cost of the port that receives the BPDU. The root path cost becomes cumulative as new BPDUs are generated.
3. Designated port election One switch port on each network segment is chosen to handle traffic for that segment. The port that announces the lowest root path cost in the segment becomes the designated port.
4. Bridging loops are removed Switch ports that are neither root ports nor designated ports are placed in the blocking state. This step breaks any bridging loops that would form otherwise.
STP Tiebreakers
When any STP decision has identical conditions or a tie, the final decision is based on this sequence of conditions:
1. The lowest BID
2. The lowest root path cost
3. The lowest sender BID
4. The lowest port ID
Path Costs
By default, switch ports have the path costs defined in Table 7-1.
Table 7-1. Switch Port Path Costs
Port Speed
Default Port Cost "Short Mode"
Default Port Cost "Long Mode"
4 mbps
250
N/A
10 mbps
100
2,000,000
16 mbps
62
N/A
45 mbps
39
N/A
100 mbps
19
200,000
155 mbps
14
N/A
622 mbps
6
N/A
1 gbps
4
20,000
10 gbps
2
2000
100 gbps
N/A
200
1000 gbps (1 tbps)
N/A
20
10 tbps
N/A
2

By default, Catalyst switches in PVST+ mode use the "short mode" or 16-bit path or port cost values. When the port speeds in a network are less than 1 gbps, the short mode scale is sufficient. If you have any ports that are 10 gbps or greater, however, set all switches in the network to use the "long mode" or 32-bit path cost scale. This ensures that root path cost calculations are consistent on all switches. Switches using MISTP, MISTP-PVST+, or MST automatically use the long-mode values.
NOTE
The IEEE uses a nonlinear scale to relate the port bandwidth of a single link to its port cost value. STP treats bundled links, such as Fast EtherChannel and Gigabit EtherChannel, as a single link with an aggregate bandwidth of the individual links. As a result, remember that the port or path cost used for a bundled EtherChannel will be based on the bundled bandwidth. For example, a two-link Fast EtherChannel has 200 mbps bandwidth and a path cost of 12. A four-link Gigabit EtherChannel has 4 gbps bandwidth and a path cost of 2. Use Table 7-1 to see how these EtherChannel aggregate bandwidth and port costs relate to the values of single or individual links.

STP Port States
Each switch port progresses through a sequence of states:
1. Disabled Ports that are administratively shut down or shut down due to a fault condition. (MST calls this state discarding.)
2. Blocking The state used after a port initializes. The port cannot receive or transmit data, cannot add MAC addresses to its address table, and can receive only BPDUs. If a bridging loop is detected, or if the port loses its root or designated port status, it will be returned to the blocking state. (MST calls this state discarding.)
3. Listening If a port can become a root or designated port, it is moved into the listening state. The port cannot receive or transmit data and cannot add MAC addresses to its address table. BPDUs can be received and sent. (MST calls this state discarding.)
4. Learning After the Forward Delay timer expires (default 15 seconds), the port enters the learning state. The port cannot transmit data, but can send and receive BPDUs. MAC addresses can now be learned and added into the address table.
5. Forwarding After another Forward Delay timer expires (default 15 seconds), the port enters the forwarding state. The port can now send and receive data, learn MAC addresses, and send and receive BPDUs.
STP Topology Changes
· If a switch port is moved into the forwarding state (except when PortFast is enabled), a topology change is signaled.
· If a switch port is moved from the forwarding or learning state into the blocking state, a topology change is signaled.
· To signal a topology change, a switch sends TCN BPDUs on its root port every hello time interval. This occurs until the TCN is acknowledged by the upstream designated bridge neighbor. Neighbors continue to relay the TCN BPDU on their root ports until it is received by the root bridge.
· The root bridge informs the entire spanning tree of the topology change by sending a configuration BPDU with the topology change (TC) bit set. This causes all downstream switches to reduce their Address Table Aging timers from the default value (300 seconds) down to the Forward Delay (default 15 seconds). This flushes inactive MAC addresses out of the table faster than normal.
Improving STP Stability
· STP Root Guard can be used to help enforce the root bridge placement and identity in a switched network. When enabled on a port, Root Guard disables the port if a better BPDU is received. This prevents other unplanned switches from becoming the root.
· STP Root Guard should be enabled on all ports where the root bridge should not appear. This preserves the current choice of the primary and secondary root bridges.
· Unidirectional Link Detection (UDLD) provides a means to detect a link that is transmitting in only one direction, enabling you to prevent bridging loops and traffic black holes that are not normally detected or prevented by STP.
· UDLD operates at Layer 2, by sending packets containing the device and port ID to connected neighbors on switch ports. As well, any UDLD packets received from a neighbor are reflected back so that the neighbor can see it has been recognized. UDLD messages are sent at the message interval times, usually defaulting to 15 seconds.
· UDLD operates in two modes:
- Normal mode Unidirectional links are detected and reported as an error, but no other action is taken.
- Aggressive mode Unidirectional links are detected, reported as an error, and disabled after eight attempts (once a second for eight seconds) to reestablish the link. Disabled ports must be manually reenabled.
· STP Loop Guard detects the absence of BPDUs on the root and alternate root ports. Nondesignated ports are temporarily disabled, preventing them from becoming designated ports and moving into the forwarding state.
· STP Loop Guard should be enabled on the root and alternate root ports (both non-designated) for all possible active STP topologies.

6-6. Private VLANs

0 comments
· Private VLANs allow for additional security between devices in a common subnet.
· Private edge VLANs can be configured to prevent connectivity between devices on access switches.
· Private VLANs can be configured on the Catalyst 6000 and Catalyst 4000 series products.
· Within a private VLAN, you can isolate devices to prevent connectivity between devices within the isolated VLAN.
· Within a private VLAN, communities can be created to allow connection between some devices and to prevent them from communicating with others.
· Promiscuous ports are mapped to private VLANs to allow for connectivity to VLANs outside of this network.

6-5. GVRP

0 comments
· Generic Attribute Registration Protocol (GARP) VLAN Registration Protocol (GVRP) is an application defined in the IEEE 802.1Q standard that allows for the control of VLANs.
· GVRP runs only on 802.1Q trunk links.
· GVRP prunes trunk links so that only active VLANs will be sent across trunk connections.
· GVRP expects to hear join messages from the switches before it will add a VLAN to the trunk.
· GVRP updates and hold timers can be altered.
· GVRP ports run in various modes to control how they will prune VLANs.
· GVRP can be configured to dynamically add and manage VLANS to the VLAN database for trunking purposes.

6-4. VLAN Trunking Protocol

0 comments
· VTP sends messages between trunked switches to maintain VLANs on these switches in order to properly trunk.
· VTP is a Cisco proprietary method of managing VLANs between switches and runs across any type of trunking mechanism.
· VTP messages are exchanged between switches within a common VTP domain.
· VTP domains must be defined or VTP disabled before a VLAN can be created.
· Exchanges of VTP information can be controlled by passwords.
· VTP manages only VLANs 2 through 1002.
· VTP allows switches to synchronize their VLANs based on a configuration revision number.
· Switches can operate in one of three VTP modes: server, transparent, or client.
· VTP can prune unneeded VLANs from trunk links.

6-3. Trunking

0 comments
· VLANs are local to each switch's database, and VLAN information is not passed between switches.
· Trunk links provide VLAN identification for frames traveling between switches.
· Cisco switches have two Ethernet trunking mechanisms: ISL and IEEE 802.1Q.
· Certain types of switches can negotiate trunk links.
· Trunks carry traffic from all VLANs to and from the switch by default but can be configured to carry only specified VLAN traffic.
· Trunk links must be configured to allow trunking on each end of the link.

6-2. VLAN Port Assignments

0 comments
·· VLANs are assigned to individual switch ports.
· Ports can be statically assigned to a single VLAN or dynamically assigned to a single VLAN.
· All ports are assigned to VLAN 1 by default.
· Ports are active only if they are assigned to VLANs that exist on the switch.
· Static port assignments are performed by the administrator and do not change unless modified by the administrator, whether the VLAN exists on the switch or not.
· Dynamic VLANs are assigned to a port based on the MAC address of the device plugged into a port.
· Dynamic VLAN configuration requires a VLAN Membership Policy Server (VMPS) client, server, and database to operate properly.

6-1. VLAN Configuration

0 comments
· VLANs are broadcast domains defined within switches to allow control of broadcast, multicast, unicast, and unknown unicast within a Layer 2 device.
· VLANs are defined on a switch in an internal database known as the VLAN Trunking Protocol (VTP) database. After a VLAN has been created, ports are assigned to the VLAN.
· VLANs are assigned numbers for identification within and between switches. Cisco switches have two ranges of VLANs, the normal range and extended range.
· VLANs have a variety of configurable parameters, including name, type, and state.
· Several VLANs are reserved, and some can be used for internal purposes within the switch.

5-6. Routing Tables

0 comments
· To move packets between separate networks, the switching processor must have knowledge of the destination network.
· Networks that are connected to a physical or virtual interface are connected routes and are automatically known by the switching processor.
· You can configure the Layer 3 switching processor with statically defined routes by entering the routes into the configuration file.
· One of the most common ways to learn and maintain routes is to use a dynamic routing protocol, such as Open Shortest Path First (OSPF) Protocol or Enhanced Interior Gateway Routing Protocol (EIGRP).

5-5. Virtual Interfaces

0 comments
· Virtual interfaces exist for configuration where there is no single physical attachment to a broadcast domain.
· For switches with Layer 2 interfaces, VLANs define broadcast domains.
· The VLAN interface is a Layer 3 interface for any member of the given VLAN.
· For switches or routers with Layer 3 interfaces, broadcast domains are defined as bridge groups.
· To route between bridge groups and other broadcast domains, a bridged virtual interface (BVI) is used as a Layer 3 interface.
· In some instances, a physical Layer 3 interface can support traffic from multiple VLANs.
· To provide Layer 3 interfaces for each VLAN on the physical connection, a sub interface is configured as the Layer 3 interface for the members of the VLAN.

5-4. WAN Interfaces

0 comments
· The Catalyst 5000 and 6000 series switches offer support for WAN interfaces to be added to the switch chassis.
· WAN interfaces are only known to the Layer 3 switching processor and must be configured from an IOS interface.
· The 5000 series switch allows for the addition of a RSM/VIP2 card, which can provide support for a variety of port adapter modules (PAMs) for WAN connectivity.
· The 6000 series switch supports a Flex WAN card, which provides support for a variety of WAN PAMs for WAN connectivity.In addition to the Flex WAN card, the 6000 series switch offers a variety of optical services modules, which can be connected to high-speed optical networks.

5-3. Layer 3 Ether Channels

0 comments
· An Ether Channel is the aggregation of multiple physical channels into a single logical connection.
· The single logical connection is referred to as a port channel.
· You can configure the port channel to operate as a Layer 3 interface on some switches.
· When assigned with an IP address, the port channel becomes the logical Layer 3 interface.
· If any single link of the channel fails, the port channel interface is still accessible through the other links.
· Layer 3 Ether Channel operation is the same as Layer 2 Ether Channels for traffic distribution and channel establishment.

5-2. Layer 3 Ethernet Interfaces

0 comments
· Layer 3 switching requires an interface on the switch that can forward packets based on Layer 3 addressing.
· Each Layer 3 interface defines a separate broadcast domain and therefore a separate network.
· After a Layer 3 interface has been configured with a protocol, it can act as a gateway for other devices in the same broadcast domain.On some switches, you can configure an Ethernet port (interface) as a Layer 3 interface