December 10, 201411 yr I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? syslog.zip
December 11, 201411 yr I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? Need more information to assist here. How much memory allocated on the host? What are your memory settings for the VM (max/min)? Are you passing through any PCI, USB, or other physical devices to the VM? Also, please post your current hardware configuration as well. VM-specific issues will require more work to track down as there are multiple places where configuration may be misaligned causing the issue to begin with (the host OS, the host hardware, the guest OS, etc).
December 11, 201411 yr Author I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? Need more information to assist here. How much memory allocated on the host? What are your memory settings for the VM (max/min)? Are you passing through any PCI, USB, or other physical devices to the VM? Also, please post your current hardware configuration as well. VM-specific issues will require more work to track down as there are multiple places where configuration may be misaligned causing the issue to begin with (the host OS, the host hardware, the guest OS, etc). First off thank you so much for taking the time to look into this. I know you guys are super busy. The dashboard is showing 3192MB memory allocated 8192 installed the memory usage graphic bar is currently at 86% but I know that when the VM is operational it is often at 95%. ( Just did a re-boot memory usage 41%, allocated 3920GB, installed 8192) My VM max and min settings are for 2GB of ram. I have to confess I am a total noob when it comes to VM's and I am not sure if these settings are right. I am not passing through any devices to the VM. VM XML <domain type='kvm'> <name>Win 8.1ProN</name> <uuid>10aad7c6-aa09-e4e0-3df2-ad918169597c</uuid> <description>HTPC Den</description> <memory unit='KiB'>2097152</memory> <currentMemory unit='KiB'>2097152</currentMemory> <vcpu placement='static'>2</vcpu> <os> <type arch='x86_64' machine='pc-i440fx-2.1'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> </features> <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/qemu-system-x86_64</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2'/> <source file='/mnt/cache/Image Media/Win8.1ProN.qcow2'/> <target dev='hda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/cache/Images/virtio-win-0.1-94.iso'/> <target dev='hdd' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='1'/> </disk> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <controller type='pci' index='0' model='pci-root'/> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <interface type='bridge'> <mac address='52:54:00:3b:8d:11'/> <source bridge='br0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0'> <listen type='address' address='0.0.0.0'/> </graphics> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain> My current hardware configuration is as follows Motherboard GA-990FXA-UD5 Revsion 3.0 Memory Kingston HyperX 8GB (2 x4GB) DDR3-1600 CL9 Cache drive is a Corsair SSD. VM and Plugins on Cache. CPU AMD FX8350 not overclocked System runs cool. Lots of fans and CPU water cooled by a coolermaster Nepton 280L (Plan to be able to run gaming/HTPC rig on top once its all working) Last night I deleted the previous Win8.1 VM and the KVM plugins and reinstalled. I used the template in the KVM manager and increased the disk size for the VM to 40GB's. No pass through. Just VNC. Windows booted, installed and updated. A few hours later it crashed or was stopped by unRaid. Syslog attached. The end of the log appears to show memory issues and processes being shutdown automatically. I am using Phazes's 64bit plugins for Sickbeard, Sabnzbd, Plex Server, CouchPotato and Headphones. It also seemed that a number of these plugins (sabnzbd and Sickbeard) also shutdown automatically when the VM went down. Is there anything else I can add that would be helpful? Thanks. syslog_1.zip
December 11, 201411 yr I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? Need more information to assist here. How much memory allocated on the host? What are your memory settings for the VM (max/min)? Are you passing through any PCI, USB, or other physical devices to the VM? Also, please post your current hardware configuration as well. VM-specific issues will require more work to track down as there are multiple places where configuration may be misaligned causing the issue to begin with (the host OS, the host hardware, the guest OS, etc). First off thank you so much for taking the time to look into this. I know you guys are super busy. The dashboard is showing 3192MB memory allocated 8192 installed the memory usage graphic bar is currently at 86% but I know that when the VM is operational it is often at 95%. ( Just did a re-boot memory usage 41%, allocated 3920GB, installed 8192) My VM max and min settings are for 2GB of ram. I have to confess I am a total noob when it comes to VM's and I am not sure if these settings are right. I am not passing through any devices to the VM. VM XML <domain type='kvm'> <name>Win 8.1ProN</name> <uuid>10aad7c6-aa09-e4e0-3df2-ad918169597c</uuid> <description>HTPC Den</description> <memory unit='KiB'>2097152</memory> <currentMemory unit='KiB'>2097152</currentMemory> <vcpu placement='static'>2</vcpu> <os> <type arch='x86_64' machine='pc-i440fx-2.1'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> </features> <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/qemu-system-x86_64</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2'/> <source file='/mnt/cache/Image Media/Win8.1ProN.qcow2'/> <target dev='hda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/cache/Images/virtio-win-0.1-94.iso'/> <target dev='hdd' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='1'/> </disk> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <controller type='pci' index='0' model='pci-root'/> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <interface type='bridge'> <mac address='52:54:00:3b:8d:11'/> <source bridge='br0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0'> <listen type='address' address='0.0.0.0'/> </graphics> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain> My current hardware configuration is as follows Motherboard GA-990FXA-UD5 Revsion 3.0 Memory Kingston HyperX 8GB (2 x4GB) DDR3-1600 CL9 Cache drive is a Corsair SSD. VM and Plugins on Cache. CPU AMD FX8350 not overclocked System runs cool. Lots of fans and CPU water cooled by a coolermaster Nepton 280L (Plan to be able to run gaming/HTPC rig on top once its all working) Last night I deleted the previous Win8.1 VM and the KVM plugins and reinstalled. I used the template in the KVM manager and increased the disk size for the VM to 40GB's. No pass through. Just VNC. Windows booted, installed and updated. A few hours later it crashed or was stopped by unRaid. Syslog attached. The end of the log appears to show memory issues and processes being shutdown automatically. I am using Phazes's 64bit plugins for Sickbeard, Sabnzbd, Plex Server, CouchPotato and Headphones. It also seemed that a number of these plugins (sabnzbd and Sickbeard) also shutdown automatically when the VM went down. Is there anything else I can add that would be helpful? Thanks. Ok, to be honest, you're in a bit of a pickle here because there could be numerous things that are causing this on your particular system. First and foremost, please disable all of your plugins and see if this behavior is recreated by using the VM by itself. If the answer is no, there is a good chance that your root cause is coming from a plugin. Second, let's try adjusting a small section on your XML to optimize for windows a bit better. Change this part of your XML: <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> to this: <clock offset='localtime'> <timer name='hypervclock'/> </clock> This may or may not assist with your issues, but it's definitely better for performance in Windows VMs to use hypervclock over the alternatives.
December 12, 201411 yr Author I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? Need more information to assist here. How much memory allocated on the host? What are your memory settings for the VM (max/min)? Are you passing through any PCI, USB, or other physical devices to the VM? Also, please post your current hardware configuration as well. VM-specific issues will require more work to track down as there are multiple places where configuration may be misaligned causing the issue to begin with (the host OS, the host hardware, the guest OS, etc). First off thank you so much for taking the time to look into this. I know you guys are super busy. The dashboard is showing 3192MB memory allocated 8192 installed the memory usage graphic bar is currently at 86% but I know that when the VM is operational it is often at 95%. ( Just did a re-boot memory usage 41%, allocated 3920GB, installed 8192) My VM max and min settings are for 2GB of ram. I have to confess I am a total noob when it comes to VM's and I am not sure if these settings are right. I am not passing through any devices to the VM. VM XML <domain type='kvm'> <name>Win 8.1ProN</name> <uuid>10aad7c6-aa09-e4e0-3df2-ad918169597c</uuid> <description>HTPC Den</description> <memory unit='KiB'>2097152</memory> <currentMemory unit='KiB'>2097152</currentMemory> <vcpu placement='static'>2</vcpu> <os> <type arch='x86_64' machine='pc-i440fx-2.1'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> </features> <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/qemu-system-x86_64</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2'/> <source file='/mnt/cache/Image Media/Win8.1ProN.qcow2'/> <target dev='hda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/cache/Images/virtio-win-0.1-94.iso'/> <target dev='hdd' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='1'/> </disk> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <controller type='pci' index='0' model='pci-root'/> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <interface type='bridge'> <mac address='52:54:00:3b:8d:11'/> <source bridge='br0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0'> <listen type='address' address='0.0.0.0'/> </graphics> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain> My current hardware configuration is as follows Motherboard GA-990FXA-UD5 Revsion 3.0 Memory Kingston HyperX 8GB (2 x4GB) DDR3-1600 CL9 Cache drive is a Corsair SSD. VM and Plugins on Cache. CPU AMD FX8350 not overclocked System runs cool. Lots of fans and CPU water cooled by a coolermaster Nepton 280L (Plan to be able to run gaming/HTPC rig on top once its all working) Last night I deleted the previous Win8.1 VM and the KVM plugins and reinstalled. I used the template in the KVM manager and increased the disk size for the VM to 40GB's. No pass through. Just VNC. Windows booted, installed and updated. A few hours later it crashed or was stopped by unRaid. Syslog attached. The end of the log appears to show memory issues and processes being shutdown automatically. I am using Phazes's 64bit plugins for Sickbeard, Sabnzbd, Plex Server, CouchPotato and Headphones. It also seemed that a number of these plugins (sabnzbd and Sickbeard) also shutdown automatically when the VM went down. Is there anything else I can add that would be helpful? Thanks. Ok, to be honest, you're in a bit of a pickle here because there could be numerous things that are causing this on your particular system. First and foremost, please disable all of your plugins and see if this behavior is recreated by using the VM by itself. If the answer is no, there is a good chance that your root cause is coming from a plugin. Second, let's try adjusting a small section on your XML to optimize for windows a bit better. Change this part of your XML: <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> to this: <clock offset='localtime'> <timer name='hypervclock'/> </clock> This may or may not assist with your issues, but it's definitely better for performance in Windows VMs to use hypervclock over the alternatives. Okay so last night I removed the plugins rebooted the server and started up the Win8 VM (suggested change was made to the XML file prior to reboot) The good news is that the VM has not crashed since it was started. CPU cycles are nice and low and with everything idling it seems nice and stable so far. The thing that I don't understand is why the memory usage is so high (94% or more) and why with 8 GB of memory installed and recognised by the system it appears that only 4GB is being used. I have attached a couple of screen grabs one of the dashboard and one of top in a telnet window. That appear to highlight this. I have carried out memtest on these two sticks of ram and there did not appear to be any problems and the system does recognise them. Is there an issue here or am I misreading this? Can I fix this problem by swapping out and or adding more ram? Should I transition to docker to get away from the plugins. I understand they are less intensive when it comes to sharing system resources. Thanks.
December 12, 201411 yr I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? Need more information to assist here. How much memory allocated on the host? What are your memory settings for the VM (max/min)? Are you passing through any PCI, USB, or other physical devices to the VM? Also, please post your current hardware configuration as well. VM-specific issues will require more work to track down as there are multiple places where configuration may be misaligned causing the issue to begin with (the host OS, the host hardware, the guest OS, etc). First off thank you so much for taking the time to look into this. I know you guys are super busy. The dashboard is showing 3192MB memory allocated 8192 installed the memory usage graphic bar is currently at 86% but I know that when the VM is operational it is often at 95%. ( Just did a re-boot memory usage 41%, allocated 3920GB, installed 8192) My VM max and min settings are for 2GB of ram. I have to confess I am a total noob when it comes to VM's and I am not sure if these settings are right. I am not passing through any devices to the VM. VM XML <domain type='kvm'> <name>Win 8.1ProN</name> <uuid>10aad7c6-aa09-e4e0-3df2-ad918169597c</uuid> <description>HTPC Den</description> <memory unit='KiB'>2097152</memory> <currentMemory unit='KiB'>2097152</currentMemory> <vcpu placement='static'>2</vcpu> <os> <type arch='x86_64' machine='pc-i440fx-2.1'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> </features> <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/qemu-system-x86_64</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2'/> <source file='/mnt/cache/Image Media/Win8.1ProN.qcow2'/> <target dev='hda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/cache/Images/virtio-win-0.1-94.iso'/> <target dev='hdd' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='1'/> </disk> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <controller type='pci' index='0' model='pci-root'/> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <interface type='bridge'> <mac address='52:54:00:3b:8d:11'/> <source bridge='br0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0'> <listen type='address' address='0.0.0.0'/> </graphics> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain> My current hardware configuration is as follows Motherboard GA-990FXA-UD5 Revsion 3.0 Memory Kingston HyperX 8GB (2 x4GB) DDR3-1600 CL9 Cache drive is a Corsair SSD. VM and Plugins on Cache. CPU AMD FX8350 not overclocked System runs cool. Lots of fans and CPU water cooled by a coolermaster Nepton 280L (Plan to be able to run gaming/HTPC rig on top once its all working) Last night I deleted the previous Win8.1 VM and the KVM plugins and reinstalled. I used the template in the KVM manager and increased the disk size for the VM to 40GB's. No pass through. Just VNC. Windows booted, installed and updated. A few hours later it crashed or was stopped by unRaid. Syslog attached. The end of the log appears to show memory issues and processes being shutdown automatically. I am using Phazes's 64bit plugins for Sickbeard, Sabnzbd, Plex Server, CouchPotato and Headphones. It also seemed that a number of these plugins (sabnzbd and Sickbeard) also shutdown automatically when the VM went down. Is there anything else I can add that would be helpful? Thanks. Ok, to be honest, you're in a bit of a pickle here because there could be numerous things that are causing this on your particular system. First and foremost, please disable all of your plugins and see if this behavior is recreated by using the VM by itself. If the answer is no, there is a good chance that your root cause is coming from a plugin. Second, let's try adjusting a small section on your XML to optimize for windows a bit better. Change this part of your XML: <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> to this: <clock offset='localtime'> <timer name='hypervclock'/> </clock> This may or may not assist with your issues, but it's definitely better for performance in Windows VMs to use hypervclock over the alternatives. Okay so last night I removed the plugins rebooted the server and started up the Win8 VM (suggested change was made to the XML file prior to reboot) The good news is that the VM has not crashed since it was started. CPU cycles are nice and low and with everything idling it seems nice and stable so far. The thing that I don't understand is why the memory usage is so high (94% or more) and why with 8 GB of memory installed and recognised by the system it appears that only 4GB is being used. I have attached a couple of screen grabs one of the dashboard and one of top in a telnet window. That appear to highlight this. I have carried out memtest on these two sticks of ram and there did not appear to be any problems and the system does recognise them. Is there an issue here or am I misreading this? Can I fix this problem by swapping out and or adding more ram? Should I transition to docker to get away from the plugins. I understand they are less intensive when it comes to sharing system resources. Thanks. Unused ram is wasted ram. Your ram is used as buffers, that will be released if the system ever need that ram for programs. Have a read over at linuxatemyram.com . As for the system only allocating 4gb, it might be a reporting issue or how the memory is allocated to vms and not the host (unraid). I'll have to take a closer look at your screen shots and attachments when I can. What does the following report from a telnet or putty session: free -m The data in top also should show the same results. To my unknowledgeable eyes it looks like only 4G is available to unraid, and about 1.3 is used for caching.
December 12, 201411 yr Author I initially posted some of these issues in the KVM section of the forum and it was suggested that I post here also. http://lime-technology.com/forum/index.php?topic=35858.240 I have been playing with setting up a Win8 VM using KVM and the Virtual Machine Management plugin. I am using the most recent Libvert support and Virtual Machine Management plugins for 6.0 Beta12. Stability - I find that the Win8.1 VM seems to crash fairly regularly and when it does it seems to effect the operation of the rest of the unRaid server. There have been a number of errors showing up regarding memory allocation in the dashboard window (see attachment). I am also attaching the syslog which shows similar errors. These errors only occur when the VM is operating, being suspended or shutdown. If I don't start the VM then the server runs error free. Most of the time once the VM crashes I cannot restart it without rebooting the entire system. The server itself generally regains full operability without a reboot but this takes some time. On one occassion I also noticed that SABnzbd threw out an error code while this was occuring Error [Errno 12] Cannot allocate memory while running par2_repair on set 222982. It has been suggested that this could be due to the VM disk being too small and I will resize this to see if that helps. Any other thoughts? Need more information to assist here. How much memory allocated on the host? What are your memory settings for the VM (max/min)? Are you passing through any PCI, USB, or other physical devices to the VM? Also, please post your current hardware configuration as well. VM-specific issues will require more work to track down as there are multiple places where configuration may be misaligned causing the issue to begin with (the host OS, the host hardware, the guest OS, etc). First off thank you so much for taking the time to look into this. I know you guys are super busy. The dashboard is showing 3192MB memory allocated 8192 installed the memory usage graphic bar is currently at 86% but I know that when the VM is operational it is often at 95%. ( Just did a re-boot memory usage 41%, allocated 3920GB, installed 8192) My VM max and min settings are for 2GB of ram. I have to confess I am a total noob when it comes to VM's and I am not sure if these settings are right. I am not passing through any devices to the VM. VM XML <domain type='kvm'> <name>Win 8.1ProN</name> <uuid>10aad7c6-aa09-e4e0-3df2-ad918169597c</uuid> <description>HTPC Den</description> <memory unit='KiB'>2097152</memory> <currentMemory unit='KiB'>2097152</currentMemory> <vcpu placement='static'>2</vcpu> <os> <type arch='x86_64' machine='pc-i440fx-2.1'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> </features> <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/qemu-system-x86_64</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2'/> <source file='/mnt/cache/Image Media/Win8.1ProN.qcow2'/> <target dev='hda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/cache/Images/virtio-win-0.1-94.iso'/> <target dev='hdd' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='1'/> </disk> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <controller type='pci' index='0' model='pci-root'/> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <interface type='bridge'> <mac address='52:54:00:3b:8d:11'/> <source bridge='br0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0'> <listen type='address' address='0.0.0.0'/> </graphics> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain> My current hardware configuration is as follows Motherboard GA-990FXA-UD5 Revsion 3.0 Memory Kingston HyperX 8GB (2 x4GB) DDR3-1600 CL9 Cache drive is a Corsair SSD. VM and Plugins on Cache. CPU AMD FX8350 not overclocked System runs cool. Lots of fans and CPU water cooled by a coolermaster Nepton 280L (Plan to be able to run gaming/HTPC rig on top once its all working) Last night I deleted the previous Win8.1 VM and the KVM plugins and reinstalled. I used the template in the KVM manager and increased the disk size for the VM to 40GB's. No pass through. Just VNC. Windows booted, installed and updated. A few hours later it crashed or was stopped by unRaid. Syslog attached. The end of the log appears to show memory issues and processes being shutdown automatically. I am using Phazes's 64bit plugins for Sickbeard, Sabnzbd, Plex Server, CouchPotato and Headphones. It also seemed that a number of these plugins (sabnzbd and Sickbeard) also shutdown automatically when the VM went down. Is there anything else I can add that would be helpful? Thanks. Ok, to be honest, you're in a bit of a pickle here because there could be numerous things that are causing this on your particular system. First and foremost, please disable all of your plugins and see if this behavior is recreated by using the VM by itself. If the answer is no, there is a good chance that your root cause is coming from a plugin. Second, let's try adjusting a small section on your XML to optimize for windows a bit better. Change this part of your XML: <clock offset='localtime'> <timer name='rtc' tickpolicy='catchup'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='yes'/> </clock> to this: <clock offset='localtime'> <timer name='hypervclock'/> </clock> This may or may not assist with your issues, but it's definitely better for performance in Windows VMs to use hypervclock over the alternatives. Okay so last night I removed the plugins rebooted the server and started up the Win8 VM (suggested change was made to the XML file prior to reboot) The good news is that the VM has not crashed since it was started. CPU cycles are nice and low and with everything idling it seems nice and stable so far. The thing that I don't understand is why the memory usage is so high (94% or more) and why with 8 GB of memory installed and recognised by the system it appears that only 4GB is being used. I have attached a couple of screen grabs one of the dashboard and one of top in a telnet window. That appear to highlight this. I have carried out memtest on these two sticks of ram and there did not appear to be any problems and the system does recognise them. Is there an issue here or am I misreading this? Can I fix this problem by swapping out and or adding more ram? Should I transition to docker to get away from the plugins. I understand they are less intensive when it comes to sharing system resources. Thanks. Unused ram is wasted ram. Your ram is used as buffers, that will be released if the system ever need that ram for programs. Have a read over at linuxatemyram.com . As for the system only allocating 4gb, it might be a reporting issue or how the memory is allocated to vms and not the host (unraid). I'll have to take a closer look at your screen shots and attachments when I can. What does the following report from a telnet or putty session: free -m The data in top also should show the same results. To my unknowledgeable eyes it looks like only 4G is available to unraid, and about 1.3 is used for caching. Thanks for the link to linuxatemyram.com I've got that now. So I did a reboot of the system with the plugins and the VM set to not come on with the server. Still shows ram at 4GB even though dashboard and info tab show 8GB installed. I have to wonder what ate the other 4GB of ram? Is there some way to reallocate it back to the pool? Once I start the VM and run free -m I can see the caching increase as it should (last screen grab). But still limited to the 4GB total ram. Is this a defect in the OS or is the memory bad? Edit. OK had a bit of a brainwave and opened up the case to check out the ram installation on the motherboard. The ram was installed in slots 2 and 3. Moved one stick so that they now occupy slots 1 and 2. Rebooted and now dashboard is showing 7952GB ram allocated. Will now reinstate plugins and VM on next reboot and see if everything will play nice. I will let you know.
December 12, 201411 yr That could definitely have been the culprit. Mobos get mad at you when you put RAM in unpaired slots like you had ;-). Let us know how things progress!
December 13, 201411 yr Author So far so good. No memory allocation errors and no crashes of the VM or any plugins. The syslog is not showing any of the errors that had been so common before. Feel a bit like I should have known better but all of the ram slots are the same colour and I never bothered to look at the superfine print on the board that says DR3_4, DR3_2, DR3_3, DR3_1 off to the side of the ram slots. Lesson learned. Won't be making that mistake again. Thanks for the help.
December 13, 201411 yr So far so good. No memory allocation errors and no crashes of the VM or any plugins. The syslog is not showing any of the errors that had been so common before. Feel a bit like I should have known better but all of the ram slots are the same colour and I never bothered to look at the superfine print on the board that says DR3_4, DR3_2, DR3_3, DR3_1 off to the side of the ram slots. Lesson learned. Won't be making that mistake again. Thanks for the help. Just to clarify, you now have a 4GB chip in slot DDR3_2 and another 4GB chip in slot DDR3_1? if so, and I believe you do, then you're now taking advantage of dual channel
December 13, 201411 yr Author So far so good. No memory allocation errors and no crashes of the VM or any plugins. The syslog is not showing any of the errors that had been so common before. Feel a bit like I should have known better but all of the ram slots are the same colour and I never bothered to look at the superfine print on the board that says DR3_4, DR3_2, DR3_3, DR3_1 off to the side of the ram slots. Lesson learned. Won't be making that mistake again. Thanks for the help. Just to clarify, you now have a 4GB chip in slot DDR3_2 and another 4GB chip in slot DDR3_1? if so, and I believe you do, then you're now taking advantage of dual channel Thats correct all working great now and no more memory allocation problems. Happy about that.
Archived
This topic is now archived and is closed to further replies.