Opened 14 years ago
Closed 9 years ago
#8476 closed defect (obsolete)
use of VBoxHeadless causes hard kernel crash on CentOS when dealing with SATA disks
Reported by: | kevin clark | Owned by: | |
---|---|---|---|
Component: | host support | Version: | VirtualBox 4.0.4 |
Keywords: | Cc: | ||
Guest type: | other | Host type: | Linux |
Description (last modified by )
When I use VBoxHeadless, I have a totally repeatable test case that causes my host operating system to completely crash. I am fairly certain that this problem has something to do with the VM's use of SATA disks, because I have been a nearly identical configuration that uses IDE disks that I have been using now for at least a month. When I migrated my process to use SATA disks, I immediately encountered this crash.
Relevant info:
The VirtualBox version I am running comes from these two files:
VirtualBox-4.0-4.0.4_70112_rhel5-1.x86_64.rpm Oracle_VM_VirtualBox_Extension_Pack-4.0.4-70112.vbox-extpack
Host OS: CentOS 5.4
uname -a
= "Linux testbox.bradfordnetworks.com 2.6.18-164.el5 #1 SMP Thu Sep 3 03:28:30 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux"
Here are the commands that I enter to reliably cause a crash:
$ ssh -X 192.168.55.66 Last login: Wed Mar 2 10:14:34 2011 from 192.168.70.80 $ $ VBoxManage unregistervm --delete 'Test Appliance' $ VBoxManage createvm --name 'Test Appliance' --ostype Linux --register Virtual machine 'Test Appliance' is created and registered. UUID: 79c68136-c2c9-4825-b5ed-4222cd2c7460 Settings file: '/home/kclark/VirtualBox VMs/Test Appliance/Test Appliance.vbox' $ VBoxManage modifyvm 'Test Appliance' --memory 768 --acpi on --boot1 dvd $ VBoxManage modifyvm 'Test Appliance' --nic1 nat $ VBoxManage modifyvm 'Test Appliance' --nic2 nat $ VBoxManage modifyvm 'Test Appliance' --ioapic on $ VBoxManage modifyvm 'Test Appliance' --cableconnected1 on --mouse ps2 --keyboard ps2 --vrde on --vrdeport 3389 --vrdeauthtype null $ VBoxManage createhd --filename test_appliance.vdi --size 50000 0%...10%...20%...30%...40%...50%...60%...70%...80%...90%...100% Disk image created. UUID: 967b4471-3016-4191-b3cc-20906d4770f3 $ VBoxManage storagectl 'Test Appliance' --name 'SATA Controller' --add sata $ VBoxManage storageattach 'Test Appliance' --storagectl 'SATA Controller' --port 0 --device 0 --type hdd --medium test_appliance.vdi $ VBoxManage storageattach 'Test Appliance' --storagectl 'SATA Controller' --port 1 --device 0 --type dvddrive --medium /path/to/some/nfs/directory/ISO/CentOS_5.4_Final.iso $ $ $ VBoxManage startvm 'Test Appliance' --type headless Waiting for the VM to power on... VM has been successfully started. [around two seconds elapse at this point] $ $ $ Write failed: Broken pipe [and now the CentOS 5.4 host system is completely crashed]
I will attach a picture I took of the console with the crash dump.
Since I have this, here is a log of me upgrading my version of VirtualBox on this CentOS 5.4 machine:
$ sudo rpm -Uvh VirtualBox-4.0-4.0.4_70112_rhel5-1.x86_64.rpm Password: warning: VirtualBox-4.0-4.0.4_70112_rhel5-1.x86_64.rpm: Header V4 DSA signature: NOKEY, key ID 98ab5139 Preparing... ########################################### [100%] 1:VirtualBox-4.0 ########################################### [100%] Creating group 'vboxusers'. VM users must be member of that group! No precompiled module for this kernel found -- trying to build one. Messages emitted during module compilation will be logged to /var/log/vbox-install.log. Stopping VirtualBox kernel modules [ OK ] Uninstalling old VirtualBox DKMS kernel modules [ OK ] Trying to register the VirtualBox kernel modules using DKMS [FAILED] (Failed, trying without DKMS) Recompiling VirtualBox kernel modules [ OK ] Starting VirtualBox kernel modules [ OK ] $
I will mention one other thing with regards to this ticket: when I reproduce this exact test with the same version of VirtualBox on a fairly vanilla x86_64 box that runs Fedora 14, everything works.
Attachments (4)
Change History (7)
by , 14 years ago
Attachment: | vbox-crash-cropped.JPG added |
---|
comment:1 by , 14 years ago
I am adding the output of /proc/cpuinfo (from the crashing host machine) in case this helps.
comment:2 by , 14 years ago
Host type: | other → Linux |
---|
Could you add the proper VBox.log file from that VM session?
comment:3 by , 9 years ago
Description: | modified (diff) |
---|---|
Resolution: | → obsolete |
Status: | new → closed |
vboxheadless crash picture