Opened 15 years ago
Closed 13 years ago
#4846 closed defect (fixed)
Virtualbox gets stuck with higher network load (Host Networking) (VERR_VMX_INVALID_VMCS_PTR guru)
Reported by: | Michael Kromer | Owned by: | |
---|---|---|---|
Component: | network/hostif | Version: | VirtualBox 3.0.4 |
Keywords: | Cc: | ||
Guest type: | Windows | Host type: | Linux |
Description
Machine Guest (Windows 2003 R2 x86) gets stuck with higher network load on host interface (intel server)
It happens sometimes with high load, not always.
VBox 3.0.4 with Guest Additions. Host is opensuse 11.1 x86_64.
Attachments (3)
Change History (14)
by , 15 years ago
comment:1 by , 15 years ago
I would like to correct ... I can reproduce problem also without a high load on the network...
The only thing I did after it was running stable for a long time: Installed Office 2003. Since that time its not possible anymore to keep that system alive for more than 5 minutes. Microsoft Update I don't even get as far as needed to select updates.
Happens also with VBoxHeadless
comment:2 by , 15 years ago
I can now tell what the problem was:
SMP. I had the machine running with 2 vCPU's. The System is a 4-Core i7 (with HT 8), setting the CPU down to one keeps the system running stable again. However the System *was* running stable before installing Office 2003, so this might be an issue to look at.
comment:3 by , 15 years ago
CPU of host
processor : 7 vendor_id : GenuineIntel cpu family : 6 model : 26 model name : Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz stepping : 4 cpu MHz : 1600.000 cache size : 8192 KB physical id : 0 siblings : 8 core id : 3 cpu cores : 4 apicid : 7 initial apicid : 7 fpu : yes fpu_exception : yes cpuid level : 11 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good xtopology tsc_reliable nonstop_tsc pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 lahf_lm ida tpr_shadow vnmi flexpriority ept vpid bogomips : 5345.96 clflush size : 64 cache_alignment : 64 address sizes : 36 bits physical, 48 bits virtual power management:
comment:4 by , 15 years ago
Summary: | Virtualbox gets stuck with higher network load (Host Networking) → Virtualbox gets stuck with higher network load (Host Networking) (VERR_VMX_INVALID_VMCS_PTR guru) |
---|
You aren't running other VM software (such as KVM) at the same time, are you?
comment:5 by , 15 years ago
No, of course not. in fact there isn't even anything installed that could collide (like kvm kernel module, xen, vmware)... host is a cleanly installed and updated opensuse 11.1 x64 with a newer kernel.
Linux borg 2.6.29.6-jen82-rt #1 SMP PREEMPT RT 2009-08-12 08:40:26 +0200 x86_64 x86_64 x86_64 GNU/Linux
by , 15 years ago
Attachment: | VBox.2.log added |
---|
crash again with only 1 CPU and 'high' network load (30m/bit steady) with hostonly interface
comment:6 by , 15 years ago
So, it seems to be the same problem again, it just takes longer with only 1 CPU
comment:7 by , 14 years ago
Found this issue and here is how it can be reproduced on FreeBSD 8.2 AMD64 host system. Following script should be launched to install virtualbox stuff on host machine:
#!/bin/tcsh echo "Executing $0..." if ($#argv != 1) then echo "Usage: $0 local_user_name" exit 1 endif set -r local_user_name = $1 set -r this_path = `dirname $0` # http://wiki.freebsd.org/VirtualBox pkg_add -r virtualbox-ose if (! -e /etc/rc.conf.virtualbox) then cp /etc/rc.conf /etc/rc.conf.virtualbox echo vboxnet_enable=\"YES\" >> /etc/rc.conf endif if (! -e /boot/loader.conf.virtualbox) then cp /boot/loader.conf /boot/loader.conf.virtualbox echo vboxdrv_load=\"YES\" >> /boot/loader.conf endif if (! -e /etc/devfs.conf.virtualbox) then cp /etc/devfs.conf /etc/devfs.conf.virtualbox echo perm cd0 0660 >> /etc/devfs.conf echo perm xpt0 0660 >> /etc/devfs.conf echo perm pass0 0660 >> /etc/devfs.conf endif pw groupmod vboxusers -m $local_user_name exit 0
This will install virtualbox 3.2.12_OSE on host machine.
After that, create FreeBSD guest machine (64-bit) and set Network to "Bridged adapter", select from list host's machine's network interface (in my case - ale0). Install guest FreeBSD VM with minimal image, open sshd there and try to rsync some big data from host machine to guest - VB crashes with VERR_VMX_INVALID_VMCS_PTR guru when network load reaches some critical peak value (usually, in 5-10 seconds). However, limiting bandwidth using rsync cmd line option (--bwlimit) to say 10 (kb) allows to avoid this crash, 50 can work up to minute, 100 crashes in ~30 sec). Also crash never happens when data transfer happens between guest machine and "internet".
So in short - crash happen when "Bridged adapter" is set to ale0 and direct data transfer happens on high speed between host and guest machines.
There is a workaround - create and use "proper" bridge (for example, like described here - http://bsdwiki.reedmedia.net/wiki/networking_qemu_virtual_bsd_systems.html). After that, when I select "bridge0" issue gone completely.
p.s. at least two other tickets may be the same issue - 4918 and 5962, just updated oldest one.
by , 14 years ago
Attachment: | ara_vb.log added |
---|
comment:8 by , 14 years ago
In previous post - After that, when I select "bridge0" issue gone completely. must be After that, when I select "tap0" issue gone completely.
comment:9 by , 14 years ago
Does the problem appear with the latest versions of VirtualBox (4.0.12 and 4.1)?
comment:10 by , 13 years ago
Problem could not be reproduced with openSUSE 11.4 (x86_64) and Virtualbox 4.1 Packages.
comment:11 by , 13 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
Good to hear. I am resolving this one then.
VirtualBox Log at crash time