Contributor
Contributor

IPv6 on Host Only Network only works intermittently if at all.

OS X 10.11.6   Fusion Pro 8.1.1 (3771013)

I've set up a custom vmnet configure as follows

Screen Shot 2016-08-11 at 12.37.29 PM.png

With ipconfig -L on the host I get the following

vmnet6: flags=8863<UP,BROADCAST,SMART,RUNNING,SIMPLEX,MULTICAST> mtu 1500

  ether 00:50:56:c0:00:06

  inet 172.16.245.1 netmask 0xffffff00 broadcast 172.16.245.255

  nd6 options=1<PERFORMNUD>


Seems odd that there is no IPv6 info but *shrug*

Static IPv6

There are two guest VMs running CentOS 7 (minimal install)

Box 1 : IPV6ADDR=fd15:4ba5:5a2b:1007::3

Box 2 : IPV6ADDR=fd15:4ba5:5a2b:1007::5

ping6 between them shows 100% packet loss, ssh times out. ssh from the host gives no route to host

Time to experiment with routing.

Added an IP to vmnet6 on the Host with 'sudo ifconfig vmnet6 inet6 fd15:4ba5:5a2b:1007::1 prefixlen 64' which gives me an ifconfig of

vmnet6: flags=8863<UP,BROADCAST,SMART,RUNNING,SIMPLEX,MULTICAST> mtu 1500

  ether 00:50:56:c0:00:06

  inet 172.16.245.1 netmask 0xffffff00 broadcast 172.16.245.255

  inet6 fe80::250:56ff:fec0:6%vmnet6 prefixlen 64 scopeid 0xe

  inet6 fd15:4ba5:5a2b:1007::1 prefixlen 64

  nd6 options=1<PERFORMNUD>

Added the default GW to the guest machines (DEFROUTE did not work). IPV6_DEFAULTGW=fd15:4ba5:5a2b:1007::1%en16777736 to the ifcfg-en16777736

Still nothing between the guest machines however the host now shows ping6 fd15:4ba5:5a2b:1007::3

PING6(56=40+8+8 bytes) fd15:4ba5:5a2b:1007::1 --> fd15:4ba5:5a2b:1007::3  and a ping6 from guest to fd15:4ba5:5a2b:1007::1 works.

So we have do have a route there now. Progress! However it is not being picked up. Lets try advertising it.

sudo sysctl -w net.inet6.ip6.use_tempaddr=1

sudo sysctl -w net.inet6.ip6.forwarding=1

sudo rtadvd -sfD vmnet6  (-f for non-daemon, D for full debug)

Pings all around! Guests can ping each other, I can ssh into them from the host and ping that GW.... for about 100 seq numbers. Then the SSH pipe breaks, pings stop and everyone is unhappy again. If I restart rtadvd I get another success for about 100 seq numbers

Things seem to break right around the prefix check...

Aug 11 13:24:21  rtadvd[6744] <Info>: <getconfig> vmnet6 isn't defined in the configuration file or the configuration file doesn't exist. Treat it as default

Aug 11 13:24:21  rtadvd[6744] <Debug>: <get_prefix> add fd15:4ba5:5a2b:1007::/64 to prefix list on vmnet6

Aug 11 13:24:21  rtadvd[6744] <Debug>: <ra_timer_update> RA timer on vmnet6 is set to 0:0

Aug 11 13:24:21  rtadvd[6744] <Debug>: <ra_timeout> RA timer on vmnet6 is expired

Aug 11 13:24:21  rtadvd[6744] <Debug>: <ra_output> send RA on vmnet6, # of waitings = 0

Aug 11 13:24:21  rtadvd[6744] <Debug>: <ra_timer_update> RA timer on vmnet6 is set to 16:0

Aug 11 13:24:21  rtadvd[6744] <Debug>: <main> set timer to 16:0. waiting for inputs or timeout

Aug 11 13:24:21  rtadvd[6744] <Debug>: <ra_input> RA received from fe80::250:56ff:fec0:6 on vmnet6

Aug 11 13:24:21  rtadvd[6744] <Debug>: <main> set timer to 15:999740. waiting for inputs or timeout

Aug 11 13:24:37  rtadvd[6744] <Debug>: <ra_timeout> RA timer on vmnet6 is expired

Aug 11 13:24:37  rtadvd[6744] <Debug>: <ra_output> send RA on vmnet6, # of waitings = 0

Aug 11 13:24:37  rtadvd[6744] <Debug>: <ra_timer_update> RA timer on vmnet6 is set to 16:0

Aug 11 13:24:37  rtadvd[6744] <Debug>: <main> set timer to 16:0. waiting for inputs or timeout

Aug 11 13:24:37  rtadvd[6744] <Debug>: <ra_input> RA received from fe80::250:56ff:fec0:6 on vmnet6

Aug 11 13:24:37  rtadvd[6744] <Debug>: <main> set timer to 15:999108. waiting for inputs or timeout

Aug 11 13:24:38  rtadvd[6744] <Debug>: <ra_input> RA received from fe80::250:56ff:fec0:2222 on vmnet6

Aug 11 13:24:38  rtadvd[6744] <Info>: <nd6_options> unknown ND option(type 24)

Aug 11 13:24:38  rtadvd[6744] <Info>: <prefix_check> preferred lifetime for fd15:4ba5:5a2b:1007::/64 inconsistent on vmnet6: 14400 from fe80::250:56ff:fec0:2222, 604800 from us

Aug 11 13:24:38  rtadvd[6744] <Info>: <prefix_check> valid lifetime for fd15:4ba5:5a2b:1007::/64 inconsistent on vmnet6: 86400 from fe80::250:56ff:fec0:2222, 2592000 from us

Aug 11 13:24:38  rtadvd[6744] <Debug>: <main> set timer to 15:295751. waiting for inputs or timeout

Anyone able to give any help in regards to this? The RA from fe80 is odd, and I have tried setting explicit pltime 14400 vltime 86400 to no effect.

Update : If I leave things alone (with rtadv running) Every once in a while about 20 pings will succeed, then start failing again)

Tags (2)
0 Kudos
4 Replies
Contributor
Contributor

I did try IPv6 autoconfig however made zero progress there,

Again two guest VMs running CentOS 7 (minimal install) set to auto configure IPv6 with IPV6_AUTOCONF and IPV6_DEFROUTE set to yes.

They receive the following IPs

Box 1 : fd15:4ba5:5a2b:1007:20c:29ff:fe56:3a1c

Box 2 : fd15:4ba5:5a2b:1007:20c:29ff:feeb:af65

ping6 between them shows 100% packet loss, ssh times out. ssh from the host gives no route to host

0 Kudos
VMware Employee
VMware Employee

Hi AdrianMulvaney74,

Thanks for reporting this. We already have bug to track this. Smiley Happy

Thanks,

Nancy

0 Kudos
Contributor
Contributor

Thanks for the quick answer Nancy, is there a known workaround or expected ETA for a fix?

Should I log a support ticket for this?

0 Kudos
Contributor
Contributor

Was there any progress on this? I'm also interested in having IPv6 networking, even if it's Host-Only.

Nancy, is there any way to follow that bug report?

0 Kudos