Opened 15 years ago
Closed 15 years ago
#5014 closed defect (fixed)
Open Solaris locks up when Fedora and XP guests are running => fixed in SVN/3.0.8
Reported by: | Jeff Hoffman | Owned by: | |
---|---|---|---|
Component: | other | Version: | VirtualBox 3.0.6 |
Keywords: | Lockup | Cc: | |
Guest type: | other | Host type: | other |
Description
I have an AMD 9500 4 core machine with 6GB memory.
OpenSolaris Host 5.11 snv_122 64 bit running Virtual Box as the Host. Fedora 11 64 bit guest operating in bridged mode, no floppy, USB off and shared files on. XP 32 bit guest operating in bridged mode, no floppy, USB off and shared files on.
After a period of time running usually 1-2 days, the mouse, keyboard and network stop responding to all OSes.
What I do know is you can not ping any of the three hosts or ssh to them. The switch from a hardware standpoint still see the OpenSolaris NIC.
From another machine I can get through the switch to my NAS device without an issue so it is not the switch.
I probably need some way of create a core file on OpenSolaris Host to catch this. Problem is if I need a keyboard to enter something that is usually locked up.
For whatever reason VirtualBox has stopped creating log files for the Guests?
Change History (18)
comment:1 by , 15 years ago
comment:3 by , 15 years ago
Well I have some crash data but it was to large to attach to the case..
Is there an FTP site where I can place it?
comment:4 by , 15 years ago
More info, the crash did lock up the video, keyboard, mouse and NIC, so I was not able to get in to run the "$C" or "$<systemdump"
I tried to change USB ports with the keyboard but no power was found on any USB port(led on keyboard did not light up)
comment:5 by , 15 years ago
Could you tell me how big the crash files are after you've compressed them?
comment:6 by , 15 years ago
376MB
One thing I noticed is this daemon is pretty active?
810 root 1 0 0 6260K 4308K cpu/0 122:17 24.55% hald-addon-netw
Current OS SunOS opensolaris 5.11 snv_122 i86pc i386 i86pc Solaris
comment:7 by , 15 years ago
Ok that's fine, could you please give me your email ID so I can tell you where to upload the file or send a mail to ramshankar dot venkataraman at sun dot com.
comment:9 by , 15 years ago
Thanks a lot for the core files. I'll investigate the issue.
I see "ifconfig e1000g0 inet6 unplumb" in the list of processes, were you running this yourself? I also see an "NOTICE: e1000g0 link is down" in the log, were you tearing down this interface explicitly while VBox was running using the interface?
comment:10 by , 15 years ago
Nope was not running ifconfig e1000g0 inet6 unplumb..
OpenSolaris also had a lot of updates so my BE is at:SunOS opensolaris 5.11 snv_123 i86pc i386 i86pc Solaris.
Still started the machine in debug mode...
comment:11 by , 15 years ago
I wonder how then we have ifconfig performing an unplumb of the e1000g0 device as well the message on the system log. Were you running any upgrade?
comment:12 by , 15 years ago
This is fixed and should be available in the next release. Thank you for the report!
comment:13 by , 15 years ago
Summary: | Open Solaris locks up when Fedora and XP guests are running. → Open Solaris locks up when Fedora and XP guests are running => fixed in SVN |
---|
comment:15 by , 15 years ago
Question 4775 says it was fixed in 3.0.4 but I am running 3.0.6. Did it reappear in 3.0.6?
comment:16 by , 15 years ago
No #4775 was reported for 3.0.4. It's only fixed > 3.0.6. You can find a test build here containing the fix: http://www.virtualbox.org/download/testcase/VirtualBox-test-SunOS-3.0.7-r52830.tar.gz (this link expires in 14 days).
comment:18 by , 15 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
Summary: | Open Solaris locks up when Fedora and XP guests are running => fixed in SVN → Open Solaris locks up when Fedora and XP guests are running => fixed in SVN/3.0.8 |
Could you try this, add to "/etc/system" these lines"
Then while booting OpenSolaris press "e" in the GRUB menu and add "-k" at the end of the boot line, like so:
Note the "-k", then while booting you will get a "Loading kernel debugger..." or similar and you should boot up normally. When the hang happens after about a minute, it should hopefully drop into kmdb and you can then do "$C" and "$<systemdump" while will create a core file in /var/crash/<host>/.