Bug 1437933
Summary: | Virt-manager fails to connect to system QEMU/KVM | ||||||
---|---|---|---|---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | František Zatloukal <fzatlouk> | ||||
Component: | virt-manager | Assignee: | Cole Robinson <crobinso> | ||||
Status: | CLOSED WORKSFORME | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||
Severity: | unspecified | Docs Contact: | |||||
Priority: | unspecified | ||||||
Version: | 30 | CC: | berrange, bugzilla, crobinso, devbattah, ibragins, nguyenkhacthanh244, rjones, virt-maint | ||||
Target Milestone: | --- | Keywords: | Reopened | ||||
Target Release: | --- | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Whiteboard: | |||||||
Fixed In Version: | Doc Type: | If docs needed, set a value | |||||
Doc Text: | Story Points: | --- | |||||
Clone Of: | Environment: | ||||||
Last Closed: | 2019-04-24 14:28:02 UTC | Type: | Bug | ||||
Regression: | --- | Mount Type: | --- | ||||
Documentation: | --- | CRM: | |||||
Verified Versions: | Category: | --- | |||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
Cloudforms Team: | --- | Target Upstream Version: | |||||
Embargoed: | |||||||
Attachments: |
|
Description
František Zatloukal
2017-03-31 13:20:43 UTC
What desktop environment are you using? Typically a polkit agent is provided by eg gnome-shell. But you can avoid the need for a password prompt by addding your user to the libvirt group: http://blog.wikichoon.com/2016/01/polkit-password-less-access-for-libvirt.html I am using gnome-shell. Shouldn't I see some window where i can type my password in case it's needed? usermod --append --groups libvirt `whoami` fixed the issue, thanks! Hmm that's weird that you are using gnome-shell but the polkit agent isn't working. Does another polkit using tool like the system-config-printer 'unlock' button launch a password prompt as expected? Yes, unlock button in system-config-printer shows pop-up asking for password. Okay well since the groups change fixed things for you, not gonna both undoing it to try and narrow what went wrong, maybe some temporary breakage. But i'll reopen this if I hear other reports I can report the same thing happened here. I was running virt-manager over a remote X11 connection, and the popup as described happened. Even better, running the usermod command fixed it. That is with virt-manager-1.4.1-2.fc25.noarch Same circumstance as Richard Jones - Was SSH'd into a headless server, running virt-manager with regular user prompted the issue. Resolved by adding user to libvirt group. Due to linkrot, I'll repost the resolution - $ usermod --append --groups libvirt `whoami` This is still a bug on Fedora 30 Workstation, clean installed from Fedora-Workstation-Live-x86_64-30-20190415.n.0.iso and updated to current u-t. libvirt-daemon-5.1.0-4.fc30.x86_64 virt-manager-2.1.0-2.fc30.noarch gnome-shell-3.32.1-1.fc30.x86_64 Details shows: Unable to connect to libvirt qemu+ssh://chris/system. authentication unavailable: no polkit agent available to authenticate action 'org.libvirt.unix.manage' Verify that the 'libvirtd' daemon is running on the remote host. Libvirt URI is: qemu+ssh://chris/system Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/connection.py", line 1012, in _do_open self._backend.open(connectauth.creds_dialog, self) File "/usr/share/virt-manager/virtinst/connection.py", line 138, in open open_flags) File "/usr/lib64/python3.7/site-packages/libvirt.py", line 104, in openAuth if ret is None:raise libvirtError('virConnectOpenAuth() failed') libvirt.libvirtError: authentication unavailable: no polkit agent available to authenticate action 'org.libvirt.unix.manage' $ sudo systemctl status libvirtd ● libvirtd.service - Virtualization daemon Loaded: loaded (/usr/lib/systemd/system/libvirtd.service; enabled; vendor preset: enabled) Active: active (running) since Sat 2019-04-20 17:03:19 MDT; 28min ago Docs: man:libvirtd(8) https://libvirt.org Main PID: 1062 (libvirtd) Tasks: 19 (limit: 32768) Memory: 56.1M CGroup: /system.slice/libvirtd.service ├─1062 /usr/sbin/libvirtd ├─1177 /usr/sbin/dnsmasq --conf-file=/var/lib/libvirt/dnsmasq/default.conf --leasefile-ro --dhcp-script=/usr/libexec/libvirt_leaseshelper └─1178 /usr/sbin/dnsmasq --conf-file=/var/lib/libvirt/dnsmasq/default.conf --leasefile-ro --dhcp-script=/usr/libexec/libvirt_leaseshelper Apr 20 17:03:22 fmac.local dnsmasq[1177]: using nameserver 75.75.76.76#53 Apr 20 17:03:24 fmac.local dnsmasq[1177]: reading /etc/resolv.conf Apr 20 17:03:24 fmac.local dnsmasq[1177]: using nameserver 75.75.75.75#53 Apr 20 17:03:24 fmac.local dnsmasq[1177]: using nameserver 75.75.76.76#53 Apr 20 17:03:24 fmac.local dnsmasq[1177]: using nameserver 2001:558:feed::1#53 Apr 20 17:03:24 fmac.local dnsmasq[1177]: using nameserver 2001:558:feed::2#53 Apr 20 17:24:24 fmac.local libvirtd[1062]: libvirt version: 5.1.0, package: 4.fc30 (Fedora Project, 2019-04-02-16:12:24, ) Apr 20 17:24:24 fmac.local libvirtd[1062]: hostname: fmac.local Apr 20 17:24:24 fmac.local libvirtd[1062]: authentication unavailable: no polkit agent available to authenticate action 'org.libvirt.unix.manage' Apr 20 17:24:24 fmac.local libvirtd[1062]: End of file while reading data: Input/output error Created attachment 1556776 [details]
journal excerpt
This is what's reported in the journal from the time I double click on the connection, enter the ssh passphrase for the test machine, and the error message from client virt-manager which really looks like it's just reporting the remote host's error.
I don't know what's going on because it fails with the same message even with a public key added to the host, and this definitely works on Fedora 29 Server because I use this exact setup all the time. Sorry, just to be clear: two laptops running Fedora 30 Workstation with the software versions listed above (they are the same). The fmac.local is the host that I want to run VM's on, that's where the errors actually occur and are recorded in the journal, and is the clean installed system. The local client I'm using virt-manager on is flap.local, is not clean installed, it was f29 upgraded to f30. And fnuc.local (not really part of this bug it's just a sanity test) is a Fedora 29 Server, to which I can successfully connect from flap.local. So yeah, there's something wrong on the clean installed F30 Workstation "fmac.local" system. But I can't figure out what it is. The "Apr 20 17:24:24 fmac.local libvirtd[1062]: End of file while reading data: Input/output error" is suspicious, an EOF? OK what file? What path? I see no SElinux AVCs. Maybe the bug is the file doesn't exist but should? Chris, thanks for the report, but in general I think it's better to file a new bug and point to an older issue than to reopen a bug that's been closed for 2 years. If it's actually bug in the code there's a good chance it's in a completely different area after two years. On fmac.local, is user 'chris' in the 'libvirt' group? Otherwise this isn't going to work, because the remote host wants to launch a polkit dialog, typically provided by desktop env like gnome-shell, but the ssh session doesn't have access to it. Try: usermod --append --groups libvirt `whoami` More info here: https://blog.wikichoon.com/2016/01/polkit-password-less-access-for-libvirt.html You may need to log out and log back in on the remote host. If that doesn't help, from the local host try: virsh --connect qemu+ssh://chris/system If that fails too then we can eliminate virt-manager from the equation >On fmac.local, is user 'chris' in the 'libvirt' group?
Nope. Fixing that fixes the problem.
Reproduced this occasionally on Fedora 36 with Gnome. (In reply to František Zatloukal from comment #2) > I am using gnome-shell. Shouldn't I see some window where i can type my > password in case it's needed? > > usermod --append --groups libvirt `whoami` > fixed the issue, thanks! it worked in my case |