** Also affects: spice-packages
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cannot start VMs without routable IPv4 address
To manage not
I opened an issue against upstream spice:
https://gitlab.freedesktop.org/spice/spice/-/issues/93
** Bug watch added: gitlab.freedesktop.org/spice/spice/-/issues #93
https://gitlab.freedesktop.org/spice/spice/-/issues/93
--
You received this bug notification because you are a member of Ubuntu
Reviewing this bug, I am not sure what we want to do here. As stated in
the first comment 13 years ago, when the spice server starts, it uses
AI_ADDRCONFIG with getaddrinfo which prevents it from binding to the
IPv4 loopback address "if the local system doesn't have at least one
IPv4 configured" (g
Thanks for bumping this bug.
Apparently we have 2 config-based workaround available (comment #9 &
#30) – thanks for those!
1/ Adding this line to /etc/libvirt/qemu.conf:
spice_listen = "::1"
2/ Adding the IPv6 loopback (::1) to the VM XML file:
To me, this sounds mostly like a fea
This is still an issue in 24.10. If a VM does not have a connection it
will still not start. ::1 in qemu.conf did not fully fix the issue in my
case. This needed to be added to the actual VM XML file under:
--
You received this bug notification because you are a member of Ubuntu
Bug
** Description changed:
qemu will not start VMs using spice or vnc displays unless there is a
routable IPv4 address on the machine, even though the error relates to
127.0.0.1
root@athens:~# ip addr
- 1: lo: mtu 65536 qdisc noqueue state UNKNOWN group
default
- link/loopback 00:00
** Tags removed: server-todo
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cannot start VMs without routable IPv4 address
To manage notifications about this bug go to:
https://bugs.l
Status changed to 'Confirmed' because the bug affects multiple users.
** Changed in: glibc (Ubuntu)
Status: New => Confirmed
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cann
** Changed in: qemu (Ubuntu)
Importance: Undecided => Medium
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cannot start VMs without routable IPv4 address
To manage notifications a
** Tags removed: qemu-21.10
** Tags added: qemu-22.04 server-todo
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cannot start VMs without routable IPv4 address
To manage notifications
Launchpad has imported 17 comments from the remote bug at
https://sourceware.org/bugzilla/show_bug.cgi?id=12398.
If you reply to an imported comment from within Launchpad, your comment
will be sent to the remote bug automatically. Read more about
Launchpad's inter-bugtracker facilities at
https://
** Also affects: glibc (Ubuntu)
Importance: Undecided
Status: New
** Bug watch added: Sourceware.org Bugzilla #12398
https://sourceware.org/bugzilla/show_bug.cgi?id=12398
** Also affects: glibc via
https://sourceware.org/bugzilla/show_bug.cgi?id=12398
Importance: Unknown
It is a pain to see how often other things beat the priority of this ...
But having a config-based (no rebuilds, no custom patches) workaround just
always de-prioritizes it :-/
One day I'll get to it ... (he said and looked at his retirement in 30
years)
** Tags removed: qemu-20.10
** Tags added
** Tags added: qemu-20.10
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cannot start VMs without routable IPv4 address
To manage notifications about this bug go to:
https://bugs.laun
Wow a bug from the stone age, so old that I never saw it before :-/
I have to thank you for the bump to it, due to the current freeze for
the coming 20.04 release I'm busy and also unlikely to b able to change
it soon.
But I have taken a todo to revisit this in depth for 20.10
I mean https://bug
I agree with @addyp95 - this is a bug in how libvirt works in IPv6-only
environments, and it is still present on focal. For posterity, the
exact details of the workaround are to add the following line to
/etc/libvirt/qemu.conf:
spice_listen = "::1"
** Changed in: qemu (Ubuntu)
Status: In
This is still a present bug. In its current defaults, the lack of a
routable ipv4 address will still cause libvirt to fail starting any
guests. While setting ::1 worked for Kevin, this doesn't fix the bug,
just works around it.
--
You received this bug notification because you are a member of Ubu
Thanks, Kevin!
** Changed in: qemu (Ubuntu)
Status: Incomplete => Invalid
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1492621
Title:
Cannot start VMs without routable IPv4 address
To mana
Given that VNC does this as well, I traced the problem all the way back
to getaddrinfo() behavior in libc. I believe this describes it:
https://bugzilla.redhat.com/show_bug.cgi?id=721350
tl;dr: It ignores loopback when enumerating addresses per family.
Workaround is to define the display as ::1, w
Attaching typescript output of requested diagnostic commands.
As OVS rather than Linux bridge is in use, I have included potentially relevant
ovs-* command output.
** Attachment added: "1492621-diag.txt"
https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1492621/+attachment/4459432/+files/14
Can you show the output of 'sudo ifconfig -a; sudo brctl show; sudo
netstat -nr' ?
Does
nc -l -4 127.0.0.1 & echo ab | nc -4 127.0.0.1
work?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1
I don't have the 'kvm' command on this machine, but it definitely
appears to be qemu related:
kjotte@athens:~$ qemu-system-x86_64 -spice addr=127.0.0.1,port=5900
(/usr/bin/qemu-system-x86_64:1785): Spice-Warning **:
reds.c:2330:reds_init_socket: getaddrinfo(127.0.0.1,5900): Address family for
ho
Thanks for reporting this bug. I suspect this is actually a libvirt
bug. To make sure, could you check whether
kvm -spice addr=127.0.0.1,port=5900
works for you? At least in an empty network namespace on my host this
seemed to work fine.
** Changed in: qemu (Ubuntu)
Status: New => Inc
23 matches
Mail list logo