NS_ERROR_FAILURE console VERR_ACCESS_DENIED

[dleclair]

Hi,

I'm not able to launch my guest Linux VM. As you can see in the snip of the file below, it dies when trying to open a socket to the console due to VERR_ACCESS_DENIED. This was working for me a few weeks ago but now I keep getting this error. I'm not sure what changed in my configuration that caused this.

I am running Virtual Box 4.3.10 on a CentOS 6.5 host with v3.14.0 Linux kernel.

When I first noticed this error, I checked /etc/group and found that my user acount was not a member of vboxusers. I added my user account to the group but it didn't make any difference to the behavior.

Any help would be appreciated.

Thanks in advance.
Denis

.
.
.
00:00:00.378207 NAT: DNS#1: 172.25.0.63
00:00:00.378609 IntNet#0: szNetwork={HostInterfaceNetworking-vboxnet0} enmTrunkType=3 szTrunk={vboxnet0} fFlags=0x8000 cbRecv=325632 cbSend=196608 fIgnoreConnectFailure=false
00:00:00.379138 Serial#0: emulating 16550A
00:00:00.379195 VMSetError: /home/vbox/vbox-4.3.10/src/VBox/Devices/Serial/DrvNamedPipe.cpp(636) int drvNamedPipeConstruct(PDMDRVINS*, CFGMNODE*, uint32_t); rc=VERR_ACCESS_DENIED
00:00:00.379198 VMSetError: NamedPipe#0 failed to bind to local socket ../dleclair_atlas_1.console
00:00:00.379231 AssertLogRel /home/vbox/vbox-4.3.10/src/VBox/Devices/Serial/DevSerial.cpp(1353) int serialConstruct(PDMDEVINS*, int, CFGMNODE*): <NULL>
00:00:00.379235 Serial0: Failed to attach to char driver. rc=VERR_ACCESS_DENIED
00:00:00.379245 PDM: Failed to construct 'serial'/0! VERR_ACCESS_DENIED (-38) - Access denied.
00:00:00.380626 AIOMgr: Flush failed with VERR_INVALID_PARAMETER, disabling async flushes
00:00:00.380700 NAT: zone(nm:mbuf_cluster, used:0)
00:00:00.380906 NAT: zone(nm:mbuf_packet, used:0)
00:00:00.380909 NAT: zone(nm:mbuf, used:0)
00:00:00.380983 NAT: zone(nm:mbuf_jumbo_pagesize, used:0)
00:00:00.381370 NAT: zone(nm:mbuf_jumbo_9k, used:0)
00:00:00.381634 NAT: zone(nm:mbuf_jumbo_16k, used:0)
00:00:00.381787 NAT: zone(nm:mbuf_ext_refcnt, used:0)
00:00:00.381820 NAT: zone(nm:mbuf_cluster, used:0)
00:00:00.382359 NAT: zone(nm:mbuf_packet, used:0)
00:00:00.382362 NAT: zone(nm:mbuf, used:0)
00:00:00.382437 NAT: zone(nm:mbuf_jumbo_pagesize, used:0)
00:00:00.382973 NAT: zone(nm:mbuf_jumbo_9k, used:0)
00:00:00.383232 NAT: zone(nm:mbuf_jumbo_16k, used:0)
00:00:00.383356 NAT: zone(nm:mbuf_ext_refcnt, used:0)
00:00:00.386072 ERROR [COM]: aRC=NS_ERROR_FAILURE (0x80004005) aIID={8ab7c520-2442-4b66-8d74-4ff1e195d2b6} aComponent={Console} aText={NamedPipe#0 failed to bind to local socket ../dleclair_atlas_1.console (VERR_ACCESS_DENIED)}, preserve=false
00:00:00.481951 Power up failed (vrc=VERR_ACCESS_DENIED, rc=NS_ERROR_FAILURE (0X80004005))

[Perryg]

The error looks like it is the serial port that you do not have access permission for. Make sure your user name has rights to ports.

[dleclair]

I'm not sure. I checked the group settings and added my user account to the 'dialout' group. That didn't change anything. It's still behaving the same way.

I have a separate workstation at the office with a similar setup and I can launch the VM over there. In fact, on that machine, my user account is not a member of either vboxusers, dialout, or any other group for that matter and it still works fine.

I'm not sure what else would prevent access to /dev/console or serial.

[Perryg]

Have you tried to run /etc/init.d/vboxdrv setup to see if you had a kernel update since the last time it worked?

[dleclair]

I wasn't aware of that command.

I tried running it and it's still doing something (hasn't returned yet). The output so far is shown here:

[dleclair@dleclair-lptp wrd]$ sudo /etc/init.d/vboxdrv setup
[sudo] password for dleclair:
Stopping VirtualBox kernel modules [ OK ]
Uninstalling old VirtualBox DKMS kernel modulesWARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_report_sync
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_power_save
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_tbl_get_from_tag
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_create_dig_out_ctls
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_parse_pin_defcfg
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol _snd_hda_set_pin_ctl
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_add_kctls
WARNING: /lib/modules/2.6.32-431.11.2.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_detect_enable
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_report_sync
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_power_save
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_tbl_get_from_tag
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_create_dig_out_ctls
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_parse_pin_defcfg
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol _snd_hda_set_pin_ctl
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_add_kctls
WARNING: /lib/modules/2.6.32-431.el6.x86_64/kernel/sound/pci/hda/snd-hda-codec-ca0132.ko needs unknown symbol snd_hda_jack_detect_enable

[dleclair]

It came back with the output below. I'm going to try reinitializing and relaunching the VM. I'll post back with my result.

Error! There are no instances of module: vboxhost
4.3.10 located in the DKMS tree.
[ OK ]
Trying to register the VirtualBox kernel modules using DKMS[ OK ]
Starting VirtualBox kernel modules [ OK ]
[dleclair@dleclair-lptp wrd]$
[dleclair@dleclair-lptp wrd]$

[dleclair]

Unfortunately, the exact same error.

[Perryg]

In the guest settings do you have any serial ports enabled?

[dleclair]

Yes there is one port enabled: Port 1. The "Port 2" is not enabled.

The Port/File Path for Port 1 is the one that the error log is complaining about not having access to.

I've attached a screen capture of the serial port settings.

[Perryg]

It is showing an invalid selection warning. What happens if you take the check mark out of the enable box?

You may have an issue with the named pipe and should look into recreating it if you can boot the guest without the serial port.

[dleclair]

The 'invalid settings detected' warning is only complaining that I've 9MB of video which is the minimum required for supporting fullscreen or seamless modes. It's not relevant for me since I'm running the VM headless.

I'll look into whether I can launch it with the console port disabled and recreating it.

[dleclair]

After discussing this with my colleagues at work they suggested that I kill all vbox related processes and restart the vboxdrv service. So I did a 'ps -ef | grep -i vbox' and killed all of the processes that were running. Then I ran 'sudo service vboxdrv restart' to restart the vboxdrv service. After that I was able to launch my VM successfully.