Toon posts:

Linux OSPF probleem

Pagina: 1
Acties:

Verwijderd

Topicstarter
In de volgende simpele setup, heb ik het volgende probleem. Router 1 kan alles en iedereen pingen, router 2 daarentegen kan alle interfaces op router 1 pingen maar niet door router 1 naar de default gateway (of het gehele 10.33.18.x netwerk)

Afbeeldingslocatie: http://www.wlanstore.nl/zebraospf.gif

ip_forward staat aan.

zebra.conf Node 1

hostname Node1
password xxx
enable password xxx

! Interfaces

interface lo
description loopback
ip address 127.0.0.1/8

interface ath0
description backbone
ip address 10.0.0.2/24
no shutdown
multicast

interface wlan0
description sector noord
ip address 10.33.28.1/25
no shutdown
no multicast

interface wlan1
description sector zuid
ip address 10.33.28.129/25
no shutdown
no multicast

log file /var/log/zebra/zebra.log


ospfd.conf Node 1

hostname Node1
password xxx
enable password xxx

interface lo

! Intefaces

interface ath0
ip ospf message-digest-key 1 md5 blabla
ip ospf authentication message-digest
ip ospf network broadcast

interface eth0

interface wlan0
ip ospf message-digest-key 1 md5 blabla
ip ospf authentication message-digest
ip ospf network non-broadcast

interface wlan1
ip ospf message-digest-key 1 md5 blabla
ip ospf authentication message-digest
ip ospf network non-broadcast

! ospf router

router ospf
ospf router-id 10.33.28.1

passive-interface lo
passive-interface eth0

! networks

!ath0
network 10.0.0.0/24 area 0

!wlan0
network 10.33.28.0/25 area 1

!wlan1
network 10.33.28.128/25 area 1

log file /var/log/zebra/ospfd.log

zebra.conf Node 2

hostname Node2
password xxx
enable password xxx

! Interfaces

interface lo
description loopback
ip address 127.0.0.1/8

interface ath0
description backbone
ip address 10.0.0.3/24
multicast
no shutdown

interface wlan0
description wlan0
ip address 10.33.15.1/24
no multicast
no shutdown

log file /var/log/zebra/zebra.log


ospfd.conf Node 2

hostname Node2
password xxx
enable password xxx

interface lo

! Interfaces

interface ath0
ip ospf message-digest-key 1 md5 blabla
ip ospf authentication message-digest
ip ospf network broadcast

interface eth0

interface wlan0
ip ospf message-digest-key 1 md5 blabla
ip ospf authentication message-digest
ip ospf network non-broadcast

! ospf router

router ospf
ospf router-id 10.33.15.1

passive-interface lo
passive-interface eth0

! networks

!ath0
network 10.0.0.0/24 area 0

!wlan0
network 10.33.15.0/24 area 2

log file /var/log/zebra/ospfd.log

routing table router 1

code:
1
2
3
4
5
6
7
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
10.33.28.0      *               255.255.255.128 U     0      0        0 wlan0
10.33.28.128    *               255.255.255.128 U     0      0        0 wlan1
10.0.0.0        *               255.255.255.0   U     0      0        0 ath0
10.33.18.0      *               255.255.255.0   U     0      0        0 eth0
10.33.15.0      10.0.0.3        255.255.255.0   UG    20     0        0 ath0
default         10.33.18.1      0.0.0.0         UG    0      0        0 eth0


routing table router 2

code:
1
2
3
4
5
6
7
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
10.33.28.0      10.0.0.2        255.255.255.128 UG    20     0        0 ath0
10.33.28.128    10.0.0.2        255.255.255.128 UG    20     0        0 ath0
10.0.0.0        *               255.255.255.0   U     0      0        0 ath0
10.33.18.0      10.0.0.2        255.255.255.0   UG    20     0        0 ath0
10.33.15.0      *               255.255.255.0   U     0      0        0 wlan0
default         10.0.0.2        0.0.0.0         UG    20     0        0 ath0


het opvragen van de routing table van router 2 geeft enkele seconden vertraging (even als arp -a) bij router 1 wordt de routing table echter wel gelijk geprint.

ospf md5 verificatie lukt en komt tot stand, dr and backup election werkt eveneens...pingen 10.33.18.1 vanaf router 2 onmogelijk.

Iemand enig idee ?

Verwijderd

waarom perse met ospf, het lijkt mij een nodeloos ingewikkelde oplossing voor een simpel probleem. Probeer het eerst eens gewoon met statische routes zonder routing protocol. den heb je i.i.g. bevestigd of het probleem wel in ospf zit.

p.s. moet je ospf area niet identiek zijn om ze elkaars routes te laten leren ?

[ Voor 17% gewijzigd door Verwijderd op 12-12-2003 16:15 ]


  • FatalError
  • Registratie: Juni 1999
  • Laatst online: 18:15
p.s. moet je ospf area niet identiek zijn om ze elkaars routes te laten leren ?
nee, hoeft niet

If it ain't broken, tweak it! | gasloos sinds oktober 2025, hoekwoning 1978 | 10kWp PV, Panasonic K serie 7kW, Atlantic Explorer V5 270L | Tesla Model Y | 3x Victron MP2 6k5, 48kWh


Verwijderd

Mmm... de route komt wel in de routing table van node 2 terecht, maar mogelijk met het verkeerde next-hop adres: je netwerktekening is anders dan je configuratie; op de tekening heeft node 2 ip adres 10.0.0.2, en node 1 10.0.0.3, terwijl het in je config precies andersom is... hoe zijn de interfaces geconfigureerd?

Verwijderd

Dat hoeft wel, maar dat zijn ze ook. Zowel node 1 als node 2 hebben een interface in Area 0.

  • FatalError
  • Registratie: Juni 1999
  • Laatst online: 18:15
Dat hoeft wel, maar dat zijn ze ook. Zowel node 1 als node 2 hebben een interface in Area 0.
Je hebt gelijk, stom van me:)
Ben 'n beetje vermoeid;)

If it ain't broken, tweak it! | gasloos sinds oktober 2025, hoekwoning 1978 | 10kWp PV, Panasonic K serie 7kW, Atlantic Explorer V5 270L | Tesla Model Y | 3x Victron MP2 6k5, 48kWh


Verwijderd

Topicstarter
jeroenr> configs van pcmcia kaarten gaat via pcmcia-cs dmv het editen etc/pcmcia/network.opts & wireless.opts en voor pci kaart /etc/network/interfaces

overigens klopt het plaatje niet met de situatie, met router 1 wordt router 2 bedoeld en andersom :/

config files

network.opts router 1
case "$ADDRESS" in
*,*,*,00:02:6F:01:xx:xx)
INFO="Sector Zuid"
IF_PORT=""
BOOTP="n"
DHCP="n"
DHCP_HOSTNAME=""
PPPOE="n"
WHEREAMI="n"
IPADDR="10.33.28.1"
NETMASK="255.255.255.128"
NETWORK="10.33.28.0"
BROADCAST="10.33.28.127"
# GATEWAY="10.0.1.1"
DOMAIN=""
SEARCH=""
DNS_1=""
DNS_2=""
DNS_3=""
MOUNTS=""
MTU=""
IPX_FRAME=""
IPX_NETNUM=""
IPMASQ="n"
NO_CHECK=n
NO_FUSER=n
;;

*,*,*,00:02:6F:01:xx:xx)
INFO="Sector Noord"
IF_PORT=""
BOOTP="n"
DHCP="n"
DHCP_HOSTNAME=""
PPPOE="n"
WHEREAMI="n"
IPADDR="10.33.28.129"
NETMASK="255.255.255.128"
NETWORK="10.33.28.128"
BROADCAST="10.33.28.255"
# GATEWAY="10.0.1.1"
DOMAIN=""
SEARCH=""
DNS_1=""
DNS_2=""
DNS_3=""
MOUNTS=""
MTU=""
IPX_FRAME=""
IPX_NETNUM=""
IPMASQ="n"
NO_CHECK=n
NO_FUSER=n
;;

esac

wireless.opts router 1

case "$ADDRESS" in
*,*,*,00:02:6F:01:xx:xx)
INFO="Sector Zuid"
ESSID="test"
NWID=""
RATE="äuto"
MODE="master"
FREQ=""
CHANNEL="11"
SENS=""
KEY=""
RTS=""
FRAG=""
IWCONFIG=""
IWSPY=""
IWPRIV=""
;;

*,*,*,00:02:6F:01:xx:xx)
INFO="Sector Noord"
ESSID="test"
NWID=""
RATE="auto"
MODE="master"
FREQ=""
CHANNEL="6"
SENS=""
KEY=""
RTS=""
FRAG=""
IWCONFIG=""
IWSPY=""
IWPRIV=""
;;

esac

network.opts router 2

case "$ADDRESS" in
*,*,*,00:02:6F:01:xx:xx)
INFO="Sector Zuid"
IF_PORT=""
BOOTP="n"
DHCP="n"
DHCP_HOSTNAME=""
PPPOE="n"
WHEREAMI="n"
IPADDR="10.33.15.1"
NETMASK="255.255.255.0"
NETWORK="10.33.15.0"
BROADCAST="10.33.15.255"
# GATEWAY="10.0.0.1"
DOMAIN=""
SEARCH=""
DNS_1=""
DNS_2=""
DNS_3=""
MOUNTS=""
MTU=""
IPX_FRAME=""
IPX_NETNUM=""
IPMASQ="n"
NO_CHECK=n
NO_FUSER=n
;;

esac

wireless.opts router 2

case "$ADDRESS" in
*,*,*,00:02:6F:01:xx:xx)
INFO="Sector Oost"
ESSID="test"
NWID=""
RATE="äuto"
MODE="master"
FREQ=""
CHANNEL="1"
SENS=""
KEY=""
RTS=""
FRAG=""
IWCONFIG=""
IWSPY=""
IWPRIV=""
;;

esac

Is het wellicht nodig hosts toe te voegen in de routing table, dit doet ospf overigens niet maw ik weet niet of het persee nodig is. Bedoel zoiets als;
code:
1
route add -host 10.0.0.3/32 ath0


met dit als resultaat
code:
1
2
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
10.0.0.3        *               255.255.255.255 UH    0      0        0 ath0


voor de personen die zich afvroegen waarom ospf te gebruiken in deze ongelofelijk simpele opzet... het betreft een testopstelling waar uiteindelijk tientallen routers in opgenomen worden.

alleen mislukt het nu al jammerlijk

[ Voor 21% gewijzigd door Verwijderd op 12-12-2003 19:13 ]


Verwijderd

Als in het plaatje router1 en 2 omgewisseld zijn, lijkt het mij dat het probleem niet in OSPF zit, dan klopt namelijk je routing table gewoon...
Router 2:
10.33.18.0 10.0.0.2 255.255.255.0 UG 20 0 0 ath0
Is dus correct: forward traffic naar router 1

Weet je wat het source adres van de ping is?
Kan je wel vanaf router 1 pingen naar dat adres?
Wat is de output van een traceroute?
Probeer eens met "tcpdump ip proto icmp" te achterhalen of de ping wel aankomt op router 1.

Maar voor zover ik kan zien werkt OSPF goed: je routing tables worden immers gevuld met de juiste gegevens...

Verwijderd

En? Hoe staat het ermee?

Verwijderd

Ik vind het toch eigenlijk wel heel erg jammer dat je ons laat nadenken over een probleem, maar vervolgens nooit meer met de oplossing komt...
Pagina: 1