NAT: ti is null. can't do any reseting connection actions

[Quickbooks Office]

Guest aborted.

Attached log file.

[vasily Levchenko]

Could you please explain, what application work in your environment?

[aponomarenko86]

crash log

[aponomarenko86]

It's reproduced also on my system. Host is Ubuntu 12.10 64-bit and the guest is the ROSA Linux 32-bit.

The application runs wget tens of thousands times. And the box is crashed 3 hours later with the attached log and the following end message in the log:

03:07:03.908854 NAT: ti is null. can't do any reseting connection actions

The message in the /var/log/syslog:

Aug 26 18:01:46 aponomarenko kernel: [2844499.885092] NAT[14566] general protection ip:7f57db2ace08 sp:7f57a47a4bf0 error:0 in libc-2.15.so[7f57db22e000+1b5000]

Please fix this issue! Thanks in advance!

[vasily Levchenko]

Could you please collect coredump and upload it to ​ftp://ftp.oracle.com/appsdev/incoming and publish file name of the core here. Note this directory is WO.

[aponomarenko86]

I created a file ".config/apport/settings" and run VirtualBox. Waiting for a crash ...

[aponomarenko86]

I got it and uploaded: _usr_lib_virtualbox_VirtualBox.1000.crash

Please check it.

[aponomarenko86]

After this crash the start of the vm became unstable: sometimes it failed to start with these lines in the log:

...

00:00:12.548937  FE: 00 00 00
00:00:12.548944  FF: 00 00 00
00:00:12.548952 !!
00:00:12.548953 !! {vgagr}
00:00:12.548954 !!
00:00:12.548958 VGA Graphics Controller (3CF): GR index 3CE:05
00:00:12.548962  GR00:00 GR01:00 GR02:00 GR03:00 GR04:00 GR05:50 GR06:05 GR07:0F GR08:FF
00:00:12.548994 !!
00:00:12.548995 !! {vgapl}
00:00:12.548995 !!
00:00:12.548998 read mode     : 0     write mode: 0
00:00:12.549004 set/reset data: 00    S/R enable: 00
00:00:12.549009 color compare : 00    read map  : 0
00:00:12.549015 rotate        : 0     function  : 0
00:00:12.549020 don't care    : 0F    bit mask  : FF
00:00:12.549025 seq plane mask: 0F    chain-4   : on
00:00:12.549031 !!
00:00:12.549031 !! {vgasr}
00:00:12.549032 !!
00:00:12.549035 VGA Sequencer (3C5): SR index 3C4:04
00:00:12.549039  SR00:03 SR01:00 SR02:0F SR03:00 SR04:0A
00:00:12.549058 !!
00:00:12.549059 !! {vgatext}
00:00:12.549059 !!
00:00:12.549063 Not in text mode!
00:00:12.549065 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
00:00:13.549837 Changing the VM state from 'RUNNING' to 'GURU_MEDITATION'.
00:02:18.125364 Console::powerDown(): A request to power off the VM has been issued (mMachineState=Stopping, InUninit=0)
00:02:18.125730 Changing the VM state from 'GURU_MEDITATION' to 'POWERING_OFF'.

...

00:02:18.233135 ERROR [COM]: aRC=E_ACCESSDENIED (0x80070005) aIID={db7ab4ca-2a3f-4183-9243-c1208da92392} aComponent={Console} aText={The virtual machine is being powered down}, preserve=false
00:02:18.266568 AIOMgr: Flush failed with VERR_INVALID_PARAMETER, disabling async flushes
00:02:18.266800 NAT: zone(nm:mbuf_cluster, used:0)
00:02:18.267447 NAT: zone(nm:mbuf_packet, used:0)
00:02:18.267456 NAT: zone(nm:mbuf, used:0)
00:02:18.267548 NAT: zone(nm:mbuf_jumbo_pagesize, used:0)
00:02:18.268192 NAT: zone(nm:mbuf_jumbo_9k, used:0)
00:02:18.268530 NAT: zone(nm:mbuf_jumbo_16k, used:0)
00:02:18.268709 NAT: zone(nm:mbuf_ext_refcnt, used:0)
00:02:18.273190 Changing the VM state from 'DESTROYING' to 'TERMINATED'.

[aponomarenko86]

failed to start after the crash

[vasily Levchenko]

That definitely not related, please check whether your case match with #11505 or #11412 ?

[aponomarenko86]

Yes, the last problem is not related. The first one is still reproducible.

[aponomarenko86]

Was the coredump helpful? I can provide any additional info if needed.

[vasily Levchenko]

Sorry, for long response. I've received core and will investigate it at the end of the week.

[aponomarenko86]

Thank you.

[vasily Levchenko]

Could you please do the following: dump your network activity on host and in guest and together with core upload on our ftp server?

[aponomarenko86]

Uploaded crash.pcap file (340.8 Mb). But core dump has not been created this time under /var/crash for unknown reason. I will try to repeat the experiment soon.

[vasily Levchenko]

Please upload the pcaps corresponding to the crash that will gives me right mapping of socket and host/guest connections states.

[aponomarenko86]

Uploaded crash.pcap file is corresponded to the crash. I've enabled network tracing, run the vm and wait until the vm is crashed. And after the crash I've loaded the pcap file to the ftp server. What is wrong?

Thank you.

[vasily Levchenko]

Replying to aponomarenko86:

What is wrong?

Ah sorry then. Perhaps, I've got you wrong because of But core dump has not been created this time. So what the names of core, guest's and host's pcap files?

[aponomarenko86]

The name of the core dump is _usr_lib_virtualbox_VirtualBox.1000.crash. The host pcap file name is crash.pcap. How can I create the guest one? This article explains only how to create a pcap file on the host.

[vasily Levchenko]

Replying to aponomarenko86:

The name of the core dump is _usr_lib_virtualbox_VirtualBox.1000.crash. The host pcap file name is crash.pcap. How can I create the guest one? This article explains only how to create a pcap file on the host.

Oh, sorry. by host pcap I've meant capturing network traffic of host interface, e.g. with Wireshark. Network_tips says how to capture guest traffic.

[aponomarenko86]

Loaded wireshark-captured.data. The box crashed 28 Oct at 22:07.

[aeichner]

Please reopen if still relevant with a recent VirtualBox release.