OpenVZ Forum


Home » General » Support » 2.6.18-53.1.6.el5.028stab053.6PAE + ipv6 = not happy
2.6.18-53.1.6.el5.028stab053.6PAE + ipv6 = not happy [message #28050] Thu, 06 March 2008 20:25 Go to next message
broquea is currently offline  broquea
Messages: 20
Registered: December 2007
Junior Member
So I was running on 2.6.18-53.1.4.el5.028stab053.4PAE up until 2.6.18-53.1.6.el5.028stab053.6PAE had been made available. After switching over, I found that the HN and the VEs all still had IPv6 configured and running on them, however even with the default route in place, they couldn't get out. I could ping the gateway, but anything beyond was "no route to host". I rebooted back into 53.4PAE and everything worked again.

I meant to bring this up earlier after the kernel had been released but have been busy. I'm still running on 53.4PAE because IPv6 connectivity is a requirement for my system.
Re: 2.6.18-53.1.6.el5.028stab053.6PAE + ipv6 = not happy [message #28073 is a reply to message #28050] Fri, 07 March 2008 13:45 Go to previous messageGo to next message
koct9i is currently offline  koct9i
Messages: 51
Registered: February 2008
Member
ipv4 network works?

you ping gateway from VE only, or ipv6 network in HN also didn't work?

gateway on another machine?

what network card used?
any vlan used?

please check sysctl net.ipv6.conf.default.forwarding
Re: 2.6.18-53.1.6.el5.028stab053.6PAE + ipv6 = not happy [message #28088 is a reply to message #28073] Fri, 07 March 2008 22:17 Go to previous message
broquea is currently offline  broquea
Messages: 20
Registered: December 2007
Junior Member
Quote:

ipv4 network works?

Yes, otherwise I couldn't have gotten into the machine, as it is remote.

Quote:

you ping gateway from VE only, or ipv6 network in HN also didn't work?

I could ping the ipv6 gateway from the HN and VE.
I couldn't get beyond the ipv6 gateway when tying to reach www.kame.net, from either HN or VE.

Quote:

gateway on another machine?

The gateway address resides on a router upstream from the machine. The network topology is: server -> Cisco L2 switch -> Foundry Netiron router.
Obviously there is no issue with the switch or routers upstream of the server, since simply changing the kernel fixes the issue.

Quote:

what network card used?

Broadcom
eth0: Tigon3 [partno(BCM95704A6) rev 2100 PHY(5704)] (PCIX:133MHz:64-bit) 10/100/1000Base-T

Quote:

any vlan used?

No vlan used on the server. The only vlan in use is between the router and switch.

Quote:

please check sysctl net.ipv6.conf.default.forwarding

I verified that it was set correctly, however that setting shouldn't affect routing for eth0 on the HN, only the ipv6 connectivity for the VEs, correct?

[Updated on: Fri, 07 March 2008 22:18]

Report message to a moderator

Previous Topic: New Kernel 2.6.24 out
Next Topic: VPS for Oracle, requires SWAP partition
Goto Forum:
  


Current Time: Thu Mar 28 08:47:09 GMT 2024

Total time taken to generate the page: 0.01628 seconds