Practice
Download as TXT, PDF0 likes319 views
Routers, switches, hubs, and bridges are networking devices that operate at different layers of the OSI model. Routers operate at layer 3 and use logical IP addresses to route packets between networks. Switches operate at layer 2 and use MAC addresses to segment collision domains and allow for full duplex communication within a broadcast domain. Hubs operate at layer 1 and are used to connect network segments but do not segment collision domains. Bridges are software-based devices that operate at layer 2 and segment broadcast domains but have fewer ports and are slower than switches.
Practice
- 1. Routers(Multiple BD and CD)(breaks BD and Coll Domain):Path selection,packet swtiching,filtering(ACL),internetworking,Routing tab Discard Broadcasts, uses layer 3 addresess (logical Addrr (IP addr)), Switch (myultiple coll domain , one BD): works on layer 2 addrr(frames), Breaks Coll domain, Fast H/w based ,more ports,CAM table,Supports full duplex communication(hence no collision) HUB (one CD and BD): used on layer 1 basically used to connect network segments, slow, Supports HAlf Duplex Bridge : S/w based , less no of ports, slow Coll Domain : Scenario in which one device is sending all other devices have to sit back and listen , if two send coll occurs OSI: open system interconnection model: REfrence model, top down approach, Vertical approach App:GUI interface between user end application interacting with data (Telnet , FTP , TFTP) Pres:data generification, formatting and translation(ASCII) Sess:Dialog Control, Session est and terminatin, maintinig seperate sessions (Segment)Trans:Flow control , sequencing , ACK, End to End delivery,Error correction,COnnectionless and Conn Oriented(TCP , UDP) (Packet)Net:Path seletion , Layer 3 Addres((IP, ARP ,RARP)) (frame)DLL:Frame forwarding , FCS , Layer 2 Addressing , Error Detection(PPP,HDLC ,Frame Relay), Unique ident of device on LAN (Bit)PHY:Bit Sycronization , generate signals 3- Way Handshake: SYN : first segment sent to sync , SYN/ACK : User for rule estbalishment and Bi Directional comm , ACK: Rules agreed ,please start transfering the data now. Flow Control : both devices are not sending , there is buffer set Seqencing:Each data segment is numbered and can be rearranged at receiving end regardless of the order in which they are received ACK:If any segment dropped , please re-transmit. Note:::PDU: At each layer Control info is added and data is renamed with somethig called PDU(Alternate name of data on each layer) Coll Domain : Colloison occurs->JAM signal sent->Random backoff algorithm is user->Back off time is trigerred, all device sitback and have same prirotiy to transfer when the timer expires. CSMA/CD(USed in a HALF Duplex communication): Avoids from two devices sending frames at the same time to avoid collision. In a full duplex transm no collison occurs all devices can send and receive.(All SWTICH Ports are full duplex) MAC Addr : ::OUI(Given by IEEE)(24 bits)::Vendor Assigned(24 bits) Devices : (Switch ,HUB) & (router,PC,Servers) Cables : Straight THr: Disimilar devices , Cross Over : Similar Devices ,Rolled Cable:Connect HOST To routers Con serial port(COM) DATA ENCAPSULATION:::: Uppr Layer data->Transport layer breaks it into segments and (adds seq nos., src & Dest Port nos, Dest IP received from upper lyr) Network Layer : Matches the destination IP addrr and Subnet mask to identify is the request for local n/w or remote n/w if local n/w ARPs for the destination MAC Addrr , else ARPs for Default gateway, It also adds TCP/UDP protocal field to ressemable the data the receiving end DLL: It is used to add source MAC , FCS to reassemable the data at receiver and place the frame on medium Routing protocol ,PPP , frame relay or these things are used to carry the information to remote network Once it reached the default gateway of the destination: At Receiver De-Capsulation:: Bits are reassembled into FRAMES , FCS is run , Destination IP and Mask is checked,ARP is sent to check for the dest host, At each layer the lower layer control information is teared of and data finally reached the host
- 2. DOD model: Application Layer, H2HOST Layer , Internet Layer, Network Access Layer Port nos: FTP:21 Telnet:23 SMTP:25 DNS: 53 DHCP: 67 TFTP: 69 HTTP: 80 HTTPS: 443 ssH: CISCO IOS: POwering on a CISCO Swtich or a Router: 1.POST: Power on self test to check all the connections and health of the switch 2.IOS Loading: The IOS is loaded from the flash memory (conf reg: 0x21002) and expanded into the RAM 3.IOS looks for the startup config in the NVRAM , if found copies into the RAM and names it as running configuration, else enters the set-up mode.(Load procedure for IOS: 1.Flash 2.TFTP server 3.ROM(Mini-ios)) Router or Switch modes: 1.User Mode(Mode>) 2. Priveleged or EXEC mode(Enable) (Mode#) 3. Global Configuration Mode(Confgure terminal command)(Mode(config)) Various prompts: 1. Interface prompt(config-if)# 2. Line Prompt(Config- line)#(for VTY Lines) 3.Routing mode: (Config-router)# CISCO command: Sh history,Hostname,Banner Motd "Delemiting character",login,password,Logging Syncronous: typing between console messsages,Do Show,Description sh run,sh startup-config,sh ip interface brief,sh interface,sh version,No shutdown(Negate command),Sh controllers s0/0/0(DCE/DTE), copy run start, erase start & reload,sh protocols ,sh dhcp bindings,sh ip route,sh processess cpu,boot system PAssword: VTY Password, enable & enable secret(type 5 encypt) password , Console , Service PAssword encrytion(type 7 encrypt): manually encrpyt all the passwords Telnet : sends password in Plain text , SSH: Sends password in encrypted format. (1.Ip domain-name xx.com , 2. username and password , 3.crypto key generate RSA and now strength of encryption , 4.now enable ssh with ip ssh version 2 and select the type of connection: 6. Transport input ssh telnet) Cisco discovery protocol(CDP): To view the directly connectd devices to a cisco router, Cisco propreitary, Uses multicast messages,updates sent every 60 secs Commands:Sh cdp neighbors,sh cdp timers,Sh cdp neighbors detail Counter for CDP is LLDP(Link layer discovery protocol): It is non cisco propreitary will give details of neighbors which are non cisco. DHCP(DORA Process) :1.DHCPClinet Brodacast l-2 and l-3 DHCPDiscover message,2.DHCP server send a layer 2 unicast DHCPOffer message for syncronization and rules, 3.DHCPClient broadcasts a DHCPRequest for IP and other relevant info, DHCP server sends a DHCPAcknowlegement and configures the client. DHCP Conflict: Two hosts having same IP Adrr:resolved via : Ping by DHCP server before assigning IP and 2. Grartious ARP Brodacst by client whem he has a IP. DHCP forwarder or Relay: Used by clients to aquire IP addreses from DHCP server when the server is located in some other LAN.(ip Helper-address DHCP Server addrr) It basically uses an interface to route all the traffic to the helper address and get all the information for clients. Componenst of cisco router: ROM(Bootstrap,mini-ios,POST), RAM(running-config, IOS, startupconfig expanded here), NVRAM(startupconfig &configurtion register) Conf Reg(value check sh ver)(used for password recovery): 0x2100:Boot to Rom- Monitor mode,2101:Boot mini-IOS from ROM, 2102:Def value(check the boot seq in NVRAM) Bypass the configuration in RAM(0x2142) and reload router say no to set up mode. copy your startup config to running config and re-enable interfaces,change password!!
- 3. Routing protocols: Purpose to keep consistency between routing tables,determine the best path within the inter-n/w(pkt delivery b/w ntws),RIPv2, EIGRP,OSPF Minimum factors for routers to route: Dest IP,Neighbor routers,Possible and best route to remote nws. If network is not directly connected to the router: 1. Dyanmic routing used to learn n/ws with the help of routing protocols running on both the routers, routers xchange info b/w each other and update each others routing table with updates, Pros :very useful in largre n/ws, less of knowledge required Cons: More processing for router,more bandwidth,more expensive router required Syntax: ip routing, router rip , version 2 , network x.x.x.x(should be a clasful network) 2. Static routing:n/ws are manually mapped by the n/w administrator. Pros:Less processing overhead cheap router can be user,more security,less bandwidth used Cons:Vault tight knowledge for admin, large networks mapping diffcult,change on one router is to be made on each router manually Syntax: IP ROUTE (Dest Network) (Subnet Mask) (Next HOP Ip Addr), If next hop router is down, The route is removed from Routing table bu remains in Runnig config 3. Default routing(Gateway of last resort): IP ROUTE 0.0.0.0 0.0.0.0 Next Hop Address :::Advertising default route: default-information originate Advertised Distance: Connceted(0)(Includes interface mareked as next hop),Static(1),EIGRP(90),OSPF(110),RIP(120),External EIGRP(170),Unnknown(255) Request timed out: Packet(ICMP reply) gets dropped to the destination nw because of an unknown error. Destination Unreachable: Router(Default Gateway) does not have an entry for the remote n/w for which the packet is destined and hence it is dropped. Distance vector routing protocols: RIP(Routing information protocol):Is a Classful routing protocol, does not Support VLSM , does not advertise MAsk,Uses Hop count as metric(max 15), Uses brodcast messages,Perdioic updates sent(every 30sec) , Sends entire routing table in each update,uses lot of bandwith.Bellman ford algo used,Has Looping issues,slow convergence. Link State routing protocol: Classless routing protocol,send bounded and trigeered upadtes,Suppport VLSM, Advertises mask in updates,uses muticast messages,Uses bandwith as metric,just send the changes in each update not the entire routing table,Less bandwidth used,DJikstra algo used,No looping issues,fast convergence. OSPF(Open shortest path first)(Pure link state):Open standard protocol mutivendor support,VLSM & IPv6 Support,Sends Multicast updates on topology change,supports discontigonous net,Hierrachical design(confine's fault tolerance to single area,reduces routing overhead,limit update broadcast),Bandwidth metric,,Djikstra algo used. OSPF Terms: Backbone area:(Area to which all routers are connected),ABR(Area Broder router):Has one more intf in multiple areas & one in BBAREA,Supports Summarization. ASBR(Autnomous system boundary router): user connect ther hetrogenous system(EIGPR,RIP etc) to OSPF area ,Summarization supported. Hello Protocol:Used for neighbor discovery,mantaining neignbor relationship,and adjceny formation,Multicasted every 10 sec at 224.0.0.5. LSA(link State advertisments):used for xchanging routing information and formation of topology table to facilitate the formation of routing table via SPF Algo. ROuter ID: basically the name of the router: Highest loopback addrr or Highest physical addrr of and interface or new router-id command,also used in election of DR&BDR
- 4. OSPF Process initiation: 1.Hello and adjacency formation 2.LSA floofing(LSU update and LSA xchange)(topology,neighbors,routing table exist) 3.SPF Calculation configuring OSPF: router ospf proces_ID , network x.x.x.x wildcard mask area NO Paramters to match for forming OSPF neighbors relationship: 1.AREA ID 2.Stub are flag 3.Netmask 4.Hello and dead time intervals 5.Authentication Password (Above factors also cause the OSPF process to malfunction ,they are the cause of miconfigured OSPF) Loopback interfaces: Used for verifying(as they remain up and active always) and Naming the routers(works as router ID) involved in router proces. OSPF verification command: show ip ospf, *sh ip ospf neighbors, *sh ip ospf intf , sh up ospf database. Multi-Area OSPF Advantages: 1.Less "memory overhead" as the topology table is small for each router which is confined to the area in which it is. 2.Less"CPU Proessing" as the tables grow small and changes in topology are confined between the routers in the same area.3.Small routing tables due to summary routes advertised by ABR. Adjancecy formation: 1.Two way comm.(xchange of hello pkts) 2.xchange of (a.DD b.LSR c.LSA between the routers) Type of link state Advts: Type-1(Router link adv.): These are xhanged within an area between the internal routers as they update their topology changes with each other 2.Type II(Network link adv): Generated by DRs to advertise the info about the router under his governance. contain: DR and BDR IP info. 3.Type-III(Summary link Adv) They are generated by ABR to xhcnage the summary routes between areas to the backbone area(##any router with partial conn to BB area is called internal router). 4.Type -iV:Basically used to inform the other routers about the path how to reach to ASBR. 5.Type-5:(ext link adv): sent by ASBRs to advertise external n/w routes. Neigbor states: DOWN:NO "Hello Pkts" observed -->INIT: Hello pkts are received but without the RID,two comm not possible.-->2WAY:RID received,BIdirectional comm start -->3.EXSTART:DR and BDR establish adjanceny and master slave relationship.(Large RID is master(aka DR))-->EXCHANGE: router starts xchanging DD,LSU.-->LOADING:LSU are sent to gather any missed info.-->FULLL: LSA are synch and adjacency established. EIGRP:Classless,Support for VLSM,Route summaries & discontigous n/ws support,advertises netmask in updates, sends bounded updates(link states prop),does not send LsA as OSPF,sends traditional updates as sent in Distance vector,Send relaiable multicast,Less bandwidth,DUAL algorith for best path selection,equal cost load balancing Paramters to match for EIGRP neigbor relationship: AS nos must match,Hello and Dead timers,Identical K Values.(Hello every 5 sec missing 3 of them means neigbot dead) Advertised distance:Remote network distance from the neighbor which is advertising the route. Feasible distance: Lowest metric path to a remote netwrok hence is stored in routing table.(Adminsitrative distance + Advertised distance by neighbor router) Tables in EIGRP:Neigbor table: Populated at the very begging while xchange of hello packets and is stored in RAM. Topology Table:(Populated via DUAL) A table which contains all the possible routes to each remote network (including best routes).These best routes are then copied into routing table and are called "successor route". Path with a
- 5. relatively higher metric is termed a "feasible succesor"(Advertised distance < Feasible distance)(prevent loops) aka backup & is stored in topology table.Note::::::Routing table: Successor whereas Topology table : Succesor and feasible successor. Relaible trasport protocol:(uses both muticasts and unicast for EIGRP runin routers)(224.0.0.10):For each neighbor a multicast is sent if the neigbor does not reply a 16 Unicasts are sent,if the neigbor then also does not reply , neighbor is considered dead.This is called """""""Reliable Multicast""""' DUAL(DIffused update algorithm):EIGPR uses this algorightm in combo with (Topology table + RTP + Hello Packets) to select & maintain best routes for a remote n/w. Dual Tasks: Dynamic route recovries , Back-up route maintanece if available , Query neigbor for unknown & alternate routes, facilitate VLSM support. Helps EIGRP to converge really FAST: EIGRP routers refer to neighbor routers to calculate their own cost to a remote n/w.If a Best route goes down it quickly scans the topology table for all the backup routes and places it in Routing table, if no back up route is found it queries to its neigbor to help out in finding an altenate path.(This process of leveraging on the other routers for path accounts for the diffusing character).EIGRP can be configured in both RIP and OSPF style via WD mask. EIGRP uses 5 matrices to compute the best path: Bandwidth(k1)(1) Delay(K2)(0) Load(K3)(1) Reliability(K4)(0) MTU(K5)(0).Higher the metric the less desireable the route Only bandwidth & delay are static and can be configured manually rest all are dynamic. One can assign weight to the metric which he wants to be the kingmaker. Once needs to assign these weights carefully as they can lead to instability and convergence problems(particularly delay and reliability). Supports load balancing by the variance command. Variance =1 equal cost load balancing , if V=2 unequal cost load balancing. EIGRP commands and TBshoting: sh ip eigrp neighbors, sh ip eigrp interfaces,sh ip eigrp topology(for sucesor routes),sh ip eigrp traffic. Mismatch k values , ACL is present or not,AS must match,Passive interfaces,Hello and dead timers,automatic summary may be causing problem,eigrp process is enabled or not. Layer2 switching: Layer 2 switch never modifies the information in the frame it simply forwards ,floods or drop it. When the switch is first powered on it learns adrr hence the switch is fast than router,hubs etc. Switch functions: 1.Address learning: It learns and stores all the mac addrr(in CAM) from which it has received frame. 2.Frame filtering/forwarding(done on CAM table basis):initally it floods the frame out all ports other than on which it has received it and later on once it has the entry in table it just forwards that frame only that port. 3.Loop Avoidance: accomplished via STP,required to overcome n/w failures.can lead to multiple frame copies , mac table thrashing and broadcast storms. Port Security: Used to restricy unauthorized access to switch ports which are not in use.Aka limit no of MAC address allowed to access the switch. Commands: switchport port-security , swport port security mode access , swtport port-sec maximun(1,2 etc),violation(restrict,protect,shutdown),swtport port-sec mac adrr or sticky,sh ip arp , show por-sec intf f0/0,sh mac addrr table. VLAN:Basically used to break brodcast domains(Bscsts control) in a flat layer2 n/w. Each vlan is in its own subnet,Hosts between the vlans cannot communicate to each other(security),Provides scalability with the help of trunk links,provides flexibility easy addtion and removal of hosts. Ports : Access ports: can be a part of only one VLAN ,unware about the TAG information associated with a frame,by default all ports are in VLAN1 native vlan Trunk ports: Used to carry multi vlan information,can be a part of more than one vlan,carries both tagged and untagged traffic.
- 6. Encapsulation types: ISL(Inet swtch link):CISCO propretary,adds control inforamtion to frame ,can only be used with FEth and Gigabith eth links,old and no used. 802.1Q Trunkin:IEEE stnd,Adds a 12 bit VLANID to the frame to make the tag diffrentiable, ****All traffic for NATIVE VLAN is untagged. Routing between VLANS:1.Each interface of the switch functions as a DG for the VLAN,2.ROAS: Creating subinterfaces on a physical interface and make them DG for each VLAN 3.Using a layer 3 switch and making certain virtual/logical interfaces and make them the DG for resp VLANs. DTP(Dyanmic trunking protocl):used to negotiate a trunk and also the encapsulation type to used over the trunk link. Configuring VLANS: sh vlan(only acces ports),Sh intfs trunk,swtport mode access vlan 3,sh interfaces intf switchport, swtch port mode trunk,access ,ip routing swport trunk encap dot1q, to add any interface : go under the intreface and type the swport mode access, trunk command and then swport access vlan 'no' Access List: Used to filter traffic on the basis of conditional statements. Rules: Read from top to bottom, will not be read once match is found, implicit deny at end Standard List(1-99): Used to deny traffic on the basis of SIP addrr. It is to be applied closest to the destination.(Access-list <1-99> permit / deny host or x.x.x.x WDmask) Extended list(100-199): Used to deny traffic on SIP,DIP,Protocol,SPORTno,DESTPORTno. Is it to be applied closest to the source. syntax( access-list <100-199> permit / deny protool(tcp/udp/Ip) <Source IP><Dest IP>eq<Dest port no>) ==>>Only once access list can be applied per interface ,perdirection,per protocol as if 2 list applied if match found in 1st list 2nd list won't even be traveresed. Direction : Inbound list: Applied inbound to interface ,Data is not routed by thr router , aka match happens before router processing Outbound list:Applied outbound to an interface,data is processed and then is matched to the list applied. Access list should have an Ip permit any statment to allow some traffic to pass else all traffic will be denied.most specific statemens should be at top Commands: sh ip access-list,ip access list standard (list of permit and deny),sh interface,sh run,go under interface access group 1 in or access-class 2 out NAT::used to allow multiple internal hosts to connect to the internet via single or a pool of global ip addrr, adds security, conservation of global IP addr. Types of NAT:Static:One to one mapping of a global IP adrr to a local IP addrr, it is used to for hosting internal web servers. Dynamic:Allocates internal Hosts with the global IP adrr on demand from a pool of global Ip addrr. Aka for each local host to get on to the network he has to borrow a global addrr,lot of addersses required. NAT overload/PAT: Most popular form of NAt allows multiple local hosts to get to the internet with the help of a single global IP addrr, It uses Port nos to diffrentiate between the request/replies from the internet or from the host. AKA many to one Mapping configuration: ip nat inside source static local IP global IP , ip nat pool todd 172.10.10.0 172.10.10.128 255.255.255.0,ip nat source list 1 pool todd,ip nat s l-1 pool todd overload ,sh ip nat transalation, sh ip nat statistics,clear ip nat statistics , ip nat inside , ip nat outside,access-list 1 permit 10.10.10.0 255.255.255.0 SPT: It create a n/w topology and uses it to block redundant links. Root Brodge:Lowest and best bridge ID, focal poin of the network, all other switched create a lowest path to the root bridge,all ports are in forwarding state,all other switches put their ports in blocking state wrt to root bridge. BPDU: are basically probes sent every 2 sec when the switch is fist powered on to built the n/w topology and check for loops within the n/w.BPDUS contain certain paramters which are matched by switches if they match a loop is there. they also contain a bridge ID, it is of 6 bytes , 2 bytes pirority 6 bytes MAC
- 7. Addrr Bridge ID is combination of Priotiy(32768) and MAc adrr.used to elect the root bridge m forwarding and blocked ports. root port:port on a non root bridge which has the lowest path cost to the root bridge. designated port: port on a non root brige which has the lowest path cost to a network segment. blocking port: does not forward anf frames to avoid loops within the network.forwarfind port: forwards and receive frames. backup port: STP: 802.1d one STP instance per network,slow convergence. PVST+: uses extended VLAN ID to differentiate between one 802.1d instance per vlan in the n/w. comparetively fast conv and also efficeint link establisment. RSTP: used port states , dicarding : njothing is happening virtually dead, blocking : listening to BPDU but will not forward frames. Listening: BPDUs listened but no entry for mac adrr in the MAC table, preparing to forward frames learning : tries to learn the n/w topology and makes entries into the MAc addrr table , does not forward frames forwarding: state in which frames are forwrded Portfas: used to disable STP on a access port, so that it does not wait for 50 sec to come into fwding state from blocked state. DHCP addrr is assigned fast BPDU guard: uses to put a porfast mode into error disabled mode when someone tries to negotiate trunk or plugs in a switch , which may cause loops. Ethenchannel: used to buindle redundant links together so that they are not blocked by STP, increases bandwith, effuiciecy , can bundle upto 8 links,FA and Eth