Module 5 Module 7 Definitions STP Bridge ID: + bridge priority MAC address
Default priority: 32.768
Port ID: port priority + port number
Path Cost: cumulative cost to reach the root switch from a switch interface from which to calculate the cost
states of a door, inside the ' STP algorithm are:
-BLOCKING-
LISTENING
-LEARNING-DISABLED-FORWARDING
timers to remember are:
-15 sec in listening state,
-15 seconds in learning state (listening + learning Timer timer = Forward delay timer)
-20 sec (10 BPDU, max_age timer) but is the waiting time of BPDUs by the switch: if nn receive BPDUs in that time then switch recalculates stp.
-300 sec aging time: when swithc it is in the state of learning, is populating a table that shows the MAC address matches incoming packets and receiving port. This table has a flush time of 300 sec: if for example a switch detects that a particular host in 30 sec ulle not announced, then removes it from the mac this table. To avoid
-balck hole nle if CVhange Topology, Cisco has created the mechanism for Topology Change Notificvation (TCN: http://www.cisco.com/en/US/tech/tk389/tk621/technologies_tech_note09186a0080094797.shtml ref.) As soon as there is a topology change to all switches are notified and in particular the aging timer is reduced from 300 to 15 sec (forwarding delay). This will restrict the time of aging lasts for a time that is equal to max_age (20 sec) + forwarding delay (15 +15 sec)
States STP A door passes from these states: Initialization
-Da-
to blocking by a blocking or listening
disabled-by listening to or learning-disabled
from learning to forwarding or to disabled
-forwarding to disabled
States ports (the states are defined depending on how incoming packets are processed): discarding
-i packets are discarded and the switch does not learn the mac address
-learning: the packets are dropped, but the mac is stored
-forwarding: packets sent. Mac I have already been learned.
Versions the STP Several versions are introduced stp:
1) Common Spanning Tree (CST), referred to as 802.1D in which a single instance of STP runs for the entire network
2) PVST + In this case there is an instance STP for each VLAN on the network. Note, the 802.1q encapsulation supports only one instance of STP, so if several instances of STP, ISL can only be used with encapsulation protocol.
3) Rapid STP, 802.1w as shown in which there is only one instance of STP with a convergence time high. In RSTP
possimo distinguish the role of ports and port status.
Role of doors:
-root-designated alternate
-backup-
4) MST, Multiple Spanning Tree in which more 'with the same VLAN traffic requirements can be arranged in individual instances of STP.
Load Balancing using STP port priority is a layer 2 protocol used to prevent layer 2 loops in collegamneto with multiple switches.
The election of the ports (designated, blocking, non-designated) takes into account several factors, including the priority of the ports.
Playing with the priorities of the doors you get a load balancing of traffic between multiple VLANs.
For example:
Obviously we have:
Or the traffic of the four VLANs properly balanced in the two trunk link.
There are two techniques you can use to get this: stp stp port priority and cost.
STP Port Priority The scheme for the port fastethernet 0 / 1 S1:
PRIORITY
VLAN ---- ------- -----------
S1 4-2 16
3-5128 (default)
In this way, the door will be forwarding the traffic in vlan 4-2 and blocking traffic for VLANs 3-5.
S1 (config) # interface fastethernet 0 / 1
S1 (config-if) # vlan 2 spanning-tree port-priority 16
S1 (config-if) # spanning-tree vlan 4 port-priority 16 Verification priorities:
S1 # show spanning-tree interface fastethernet 0 / 1 Vlan Role Sts Cost Prio.Nbr
Type -------------- - ---- --- --------- -------- ------------------------ --------
VLAN0001 Desg FWD 19 128.1 P2p
VLAN0002 Desg FWD 19 1.16 P2p
VLAN0003 Desg FWD 19 128.1 P2p
VLAN0004 Desg FWD 19 1.16 P2p
VLAN0005 Desg FWD 19 128.1 P2p
The same is repeated for all trunk ports in order to respect the logical pattern above
S1 # show spanning-tree interface fastethernet 0 / 2 Vlan Role Sts Cost Prio.Nbr
Type ---------------- ---- --- --------- -------- - -------------------------------
VLAN0001 Desg FWD 19 128.2 P2p
VLAN0002 Desg FWD 19 128.2 P2p
VLAN0003 Desg FWD 19 16.2 P2p
VLAN0004 Desg FWD 19 128.2 P2p
VLAN0005 Desg FWD 19 2.16 P2p
Root Guard enables you to keep a detrminato root switch, even if a new switch with spade connected BID lower and therefore potentially eligible to root.
And 'Just configure root guard, for example on the interface and as soon as the BID interface receives a BPDU with the most' lower than that of the root comes in inconsistent state.
soon no longer receive 'this BPDU, the interface "through" all states and returns to the designated STP mode.
SSO, RPR +, NSF, HSRP The Catalyst switches, such as the 6500 series, are equipped with the ability to insert different modules and get a device that meets your business needs.
Supervisor Engine in some models like the 4500 there are two slots in the chassis, designed for insertion of the supervisor engine, which has the task to forward packets.
For example, the Supervisor Engine 6-E, used on the Catalyst 4500 offers a bandwidth of 320Gbps (full duplex) and 250 million packets per second.
function of the supervisor engine-> IP-forwarding
-L2, L3 MPLS
-4 10 GE ports, 48 \u200b\u200b10/100/1000 ports fasteethernet (SE 720 Data Sheet)
-have-an integrated route processor
have the control plane and data plane
supervisor engine redundancy: As the supervisor engines are important to the operation of the switch, there can not be allowed to work or have problems, then using techniques redundancy are to install two supervisor engines on the switch, so that one is always active and the other in standby. The two supervisors communicate with each other, and switch to standby comes into play when the active switch has problems.
Once turned on, both the active and standby supervisor engines, perform the bootup and "initialize" only up to certain levels, and possible redundancy of the various techniques differ with each other on the amount and type of info initialized during bootup by the supervisor engine that is on standby.
It 'clear that the greater the amount of info loaded, the less time the switchover by the swicth to stand by.
The three techniques are:
-RPR-RPR +
-SSO difference between RPR and RPR + RPR: first technique used for the redundancy of the supervisor engine.
switchover occurs:
-manual switchover
RP-or SP-crashes
you remove the active supervisor fails
-synchronization between the two supervisor
Features: startup and boot configuration registers are synchronized between the active and standby supervisors. The image of the software between the active and standby supervisor
not necessarily the same. At the time of switchover, the standby supervisor becomes active, but must complete the boot process. In addition, all line cards (ports) and hardware are rilodate reprogrammed. The switchover time is one minute.
as RPR RPR + swicthover but it is more 'quick and also the startup-configuration running-configuration and are copied to both the supervisor. The stand-by
supervisro boot has completed and the line cards do not have once been the rilodare swicthover. Before swicthover the stand-by router is synchronized. The link layer control-plane or not sicronizzati, interfaces flapped and hardware could be reprogrammed. In this case the two images must be the same. The switchover is 30 sec.
If the two supervisor engines do not have the same version of IOS the redundant supervisor works in RPR mode.
Table FIB (Forwarding Information Base) is deleted during the switchover, and then also the dynamic route, so the routing will not work properly until the route table does not riconverge, while the static routes are maintained because they are stored in configuration files is copied to the startup configuration file.
Whether you RPR RPR + switchover, which is not transparent, eg if you have a VoIP phone connected to the switch during the switchover, the call is lost.
SSO Stateful SwicthOver Improvement dele RPR +.
The goal is to have a switchover that is transparent to the end user, at least for certain types of L2 protocols and IOS.
We talk about HA-aware applications as applications synchronized between the active and stand-by. As
PFC (policy-feature-card) and DFC (Distributed-forwarding-card) hardware tables are always synchronized between the two supervisor this allows a transparent swicthover data plane to L2 and L4.
Then, synchronized data plane involves continuous packet forwarding during the failover of routing protocols against being
-aware applications has not are not synchronized. SSO-plane switchover date is 0-3 sec.
NSF Non Stop Forwarding is an interactive method that aims to quickly rebuild the RIB (Routing Information Base), which in turn allows the construction of the FIB used to perform the CEF. Through the NSF, the router that needs assistance in the reconstruction of the RIB is also in the neighboring router configured with NSF. NSF supports EIGRP, OSPF, IS-IS. Works in partnership with SSO. As soon as there is a switchover between the active RP and standby RP (Route Plan), by NSF, the interfaces remain in the UP as well as the neighboring router interfaces (NSF peer) that support NSF, so there is a continuous forwarding packets during switchover, thus avoiding the flapping interface.
It 'clear that I have to first activate and then configure SSO also NSF
High Availability: HSRP, VRRP, GLBP Typically, these types of protocols within the category of First Hop Redundancy Protocol (FHRP)
HSRP Protocolo used to have redundancy in the default gateway. The basics are: you configure an instance-
HSRP-all 'inside of that body part in one or more' router-router-priority
more 'high-
router becomes active if all routers have the same priority instead becomes active with the physical address of the router ip more 'high-
is then elected a stand-by router
fano all other routers that belong to the same instance instead remain in a state of waiting to listen and then take on the role or Active or standby router.
-all 'inside of an instance of the values \u200b\u200bof the HSRP timers "Hello" and "Standby" can be changed on a router, but the values \u200b\u200bset on the Active router overwrites the values \u200b\u200bof all other routers within the same instance
-owner is defined by RFC 2281 Cisco
% MD-Supports authentication
-you can assign group numbers ranging from 0 to 255-
with the preempt command, within a group HSRP, the router with the highest priority becomes the active router, otherwise it becomes a router that does the first boot (the priority of a router is daefult 100). Here's an example:
The R5 router has a priority of 105, while R6 100. But after doing 'before then R5 and R6, R6 is that it becomes the active router, while R5 is on standby. In the case I
preempt configured on R5, the R5 router would always be the active router, even if the boot is made after R6.
Once HSRP is configured on an interface, the router through a number of states needed to understand its role in the HSRP group. The states are:
-Disabled
-Init-Listen-Speak
-Active-Standby Virtual
The MAC has the form: 0000: OC07: acXX, where XX is the group number in hexadecimal.
Clearly in this case, the preempt command is not set in either of two routers.
Throubleshooting: For the CCNP exam, the application is handled pg.184 Whereas:
Timer Every 3 seconds hello messages are sent
Holdtwon timer is 10 sec , standby timer is 10 sec, 10 sec timer is active
Track E 'can be controlled interfaces: for example, if an interface goes down it is possible to decrease the priority according to the value of router config (default is 10). Clearly, when the interface comes back up, the priority is always increased as the configured value (10 by default).
GLBP Gateway Load Balancing Protocol is a Cisco protocol that provides load balancing.
% MD-Supports authentication
seguto of a possible scenario GLBP:
Whether the IP address that when we speak of the MAC address of virtual addresses, so:
-l 'ip virtual 10.88 1.10
-the virtual machine is assigned to the master (to review)
The default priority is 100. GLBP is Cisco proprietary.
within a GLBP group is: an active-
router (active state) that has the role of AVF and AVG
-one backup router assumes the role of stand-by AVG and AVF-
other routers in the unit, they can assume the role of AVF
-roter added in any other group were
remain in listen to the different AVF AVG assigns the virtual MAC address. As soon as a client, which uses ip as default gateway the virtual IP, it sends an ARP request, the AVG responds by assigning as the defualt gateway MAC address for the host, one of the virtual MAC address associated with an AVF (the choice of MAC is to be awarded with one of the following weighted load balancing, host-dependent, round-robin)
By default, the GLBP operations are based on host and using a round robin algorithm.
AVG is the router priority to more 'high or ip or the url higher.
Weighting Thresholds->
's a way to make an intelligent load and weighed.
Configuration Example:
track 1 interface POS 5 / 0 ip routing [link protocol]
track 1 interface POS 6 / 0 ip routing [link protocol]
interface fastethernet 0 / 0
glbp 10 weighting 110 lower 95 upper 105
glbp 10 weighting track 1 decrement 10
glbp 10 weighting track 2 decrement 10
All this indicates that:
- the two interfaces POS (Packet Over Sonet) are on track, or if it keeps an eye on their behavior. ip routing indicates that you check if the interface has an active routing protocol and IP address with a link protocol instead it keeps an eye on whether the interface is up
-In the interface which serves as a default gateway for each host, we have:
-indicates the starting value of the priority glbp, in this case 110 and two are fixed treshold: a lower bound and an upper bound.
glbp 10 weighting 110 lower 95 upper 105 gblp If the priority of the router is less than the lower bound, in this case 95, the router does not work anymore 'as the default gateway and resumes only when its priority is greater Upper bound, in this case 105.
-on 'interface and then apply two rules of track:
glbp 10 weighting track 1 decrement 10
glbp 10 weighting track 2 decrement 10 So, if the POS 5 / 0 no ip address configured, the priority is glbp decreased by 10 and reaches 90
S (config-if) # group glbp weightig maximum [lower lower] [upper upper]
glbp show vlan Command
111 reads as follows:
-
Voice Priority 150 (configured) means that the router was configured with glbp a priority of 150
-
Weighting 145 (configured 150), Thresholds: lower 95, upper 140 indicates that, because the track is configured, the priority of the router glbp now around 145! (Esmpio: HSRP mio.net)
VRRP The possible states are:
-Init-Backup-Master
Again there is the concept of virtual IP address, but unlike HSRP (Cisco-proprietary ), in VRRP, correct address as the default gateway, you can use a real ip address of a virtual interface that clearly belongs algruppo VRRP. He has a Master
and more 'backup.
% MD-Supports authentication
-The master router is the lowest priority 'high.
-VRRP does not provide mechanisms to track the interface.
-The virtual MAC address is of the form 0000.5e00.01xx where xx is the hex VRRP group. The preemption is enabled
default: the router with the highest priority 'is always high in the master VRRP group.
In the case instead of using default gateway ip address of a real, then in this case, the router that owns that address become master always regardless of priority.
Example:
In this case we have:
Router A is the AVG group 1. The client 1 or 2, as soon as the mac ip must resolve the default gateway assigned (ip virtual glbp) send an ARP request to router A which, on a round robin and at the individual request by one of the two client, or send as virtual MAC his or that of the router B. In this case, the client sends its first packet to router A and the client 2 sends it to Router B.
IRDP ICMP Router Discovery Protocol is going to, and is a protocol used by the host if they lose the connection with their default gateway. Then shall endeavor to send ICMP requests to discover a new default gateway.
-defines as the voice vlan