Sent: Wednesday, November 01, 2023 at 10:06 AM From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; To: "daggs" <daggs(a)gmx.com&gt; Cc: users(a)lists.libvirt.org Subject: Re: hdd kills vm On Tue, Oct 31, 2023 at 05:58:32PM +0100, daggs wrote: >> Sent: Thursday, October 26, 2023 at 9:50 AM >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> To: "daggs" <daggs(a)gmx.com&gt; >> Cc: libvir-list(a)redhat.com >> Subject: Re: hdd kills vm >> >> On Wed, Oct 25, 2023 at 03:06:55PM +0200, daggs wrote: >> >> Sent: Tuesday, October 24, 2023 at 5:28 PM >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> >> To: "daggs" <daggs(a)gmx.com&gt; >> >> Cc: libvir-list(a)redhat.com >> >> Subject: Re: hdd kills vm >> >> >> >> On Mon, Oct 23, 2023 at 04:59:08PM +0200, daggs wrote: >> >> >Greetings Martin, >> >> > >> >> >> Sent: Sunday, October 22, 2023 at 12:37 PM >> >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> >> >> To: "daggs" <daggs(a)gmx.com&gt; >> >> >> Cc: libvir-list(a)redhat.com >> >> >> Subject: Re: hdd kills vm >> >> >> >> >> >> On Fri, Oct 20, 2023 at 02:42:38PM +0200, daggs wrote: >> >> >> >Greetings, >> >> >> > >> >> >> >I have a windows 11 vm running on my Gentoo using libvirt (9.8.0) + qemu (8.1.2), I'm passing almost all available resources to the vm >> >> >> >(all 16 cpus, 31 out of 32 GB, nVidia gpu is pt), but the performance is not good, system lags, takes long time to boot. >> >> >> >> >> >> There are couple of things that stand out to me in your setup and I'll >> >> >> assume the host has one NUMA node with 8 cores, each with 2 threads as, >> >> >> just like you set it up in the guest XML. >> >> >thats correct, see: >> >> >$ lscpu | grep -i numa >> >> >NUMA node(s): 1 >> >> >NUMA node0 CPU(s): 0-15 >> >> > >> >> >however: >> >> >$ dmesg | grep -i numa >> >> >[ 0.003783] No NUMA configuration found >> >> > >> >> >can that be the reason? >> >> > >> >> >> >> no, this is fine, 1 NUMA node is not a NUMA, technically, so that's >> >> perfectly fine. >> >thanks for clarifying it for me >> > >> >> >> >> >> >> >> >> * When you give the guest all the CPUs the host has there is nothing >> >> >> left to run the host tasks. You might think that there "isn't >> >> >> anything running", but there is, if only your init system, the kernel >> >> >> and the QEMU which is emulating the guest. This is definitely one of >> >> >> the bottlenecks. >> >> >I've tried with 12 out of 16, same behavior. >> >> > >> >> >> >> >> >> * The pinning of vCPUs to CPUs is half-suspicious. If you are trying to >> >> >> make vCPU 0 and 1 be threads on the same core and on the host the >> >> >> threads are represented as CPUs 0 and 8, then that's fine. If that is >> >> >> just copy-pasted from somewhere, then it might not reflect the current >> >> >> situation and can be source of many scheduling issues (even once the >> >> >> above is dealt with). >> >> >I found a site that does it for you, if it is wrong, can you point me to a place I can read about it? >> >> > >> >> >> >> Just check what the topology is on the host and try to match it with the >> >> guest one. If in doubt, then try it without the pinning. >> >I can try to play with it, what I don't know is what should be the mapping logic? >> > >> >> Threads on the same core in the guest should map to threads on the same >> core in the host. Since there is no NUMA that should be enough to get >> the best performance. But even misconfiguration of this will not >> introduce lags in the system if it has 8 CPUs. So that's definitely not >> the root cause of the main problem, it just might be suboptimal. >> >> >> >> >> >> >> >> >> * I also seem to recall that Windows had some issues with systems that >> >> >> have too many cores. I'm not sure whether that was an issue with an >> >> >> edition difference or just with some older versions, or if it just did >> >> >> not show up in the task manager, but there was something that was >> >> >> fixed by using either more sockets or cores in the topology. This is >> >> >> probably not the issue for you though. >> >> >> >> >> >> >after trying a few ways to fix it, I've concluded that the issue might be related to the why the hdd is defined at the vm level. >> >> >> >here is the xml: https://bpa.st/MYTA >> >> >> >I assume that the hdd sits on the sata ctrl causing the issue but I'm not sure what is the proper way to fix it, any ideas? >> >> >> > >> >> >> >> >> >> It looks like your disk is on SATA, but I don't see why that would be an >> >> >> issue. Passing the block device to QEMU as VirtIO shouldn't cause that >> >> >> much of a difference. Try measuring the speed of the disk on the host >> >> >> and then in the VM maybe. Is that SSD or NVMe? I presume that's not >> >> >> spinning rust, is it. >> >> >as seen, I have 3 drives, 2 cdroms as sata and one hdd pt as virtio, I read somewhere that if the controller of the virtio >> >> >device is sata, than it doesn't uses the virtio optimally. >> >> >> >> Well it _might_ be slightly more beneficial to use virtio-scsi or even >> >> <disk type='block' device='lun'>, but I can't imagine that would make >> >> the system lag. I'm not that familiar with the details. >> >configure virtio-scsi and sata-scai at the same time? >> > >> >> Yes, forgot that, sorry. Try virtio-scsi. You could also go farther >> and pass through the LUN or the whole HBA (if you don't need to access >> any other disk on it) to the VM. Try the information presented here: >> >> https://libvirt.org/formatdomain.html#usb-pci-scsi-devices >> >> >> >> >> >it is a spindle, nvmes are too expensive where I live, frankly, I don't need lightning fast boot, the other BM machines running windows on spindle >> >> >run it quite fast and they aren't half as fast as this server >> >> > >> >> >> >> That might actually be related. The guest might think it is a different >> >> type of disk and use completely suboptimal scheduling. This might >> >> actually be solved by passing it as <disk device='lun'..., but at this >> >> point I'm just guessing. >> >I'll look into that, thanks. > >so bottom line, you suggest the following: >1. remove the manual cpu pin, let qemu sort that out. You might try it, of course pinning it is in the end the better option. >2. add a virtio scsi controller and connect the os hdd to it >3. pass the hss via scsi pt and not dev node >4. if I able to do #3, no need to add device='lun' as it won't use the disk option > First try (3), then you don't need to do anything else and if that succeeds you have the superior configuration. If you can pass through something that will not remove anything from your host system. >Dagg. >

Sent: Thursday, November 02, 2023 at 2:34 PM From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; To: "daggs" <daggs(a)gmx.com&gt; Cc: users(a)lists.libvirt.org Subject: Re: hdd kills vm On Wed, Nov 01, 2023 at 10:24:05PM +0100, daggs wrote: >> Sent: Wednesday, November 01, 2023 at 10:06 AM >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> To: "daggs" <daggs(a)gmx.com&gt; >> Cc: users(a)lists.libvirt.org >> Subject: Re: hdd kills vm >> >> On Tue, Oct 31, 2023 at 05:58:32PM +0100, daggs wrote: >> >> Sent: Thursday, October 26, 2023 at 9:50 AM >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> >> To: "daggs" <daggs(a)gmx.com&gt; >> >> Cc: libvir-list(a)redhat.com >> >> Subject: Re: hdd kills vm >> >> >> >> On Wed, Oct 25, 2023 at 03:06:55PM +0200, daggs wrote: >> >> >> Sent: Tuesday, October 24, 2023 at 5:28 PM >> >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> >> >> To: "daggs" <daggs(a)gmx.com&gt; >> >> >> Cc: libvir-list(a)redhat.com >> >> >> Subject: Re: hdd kills vm >> >> >> >> >> >> On Mon, Oct 23, 2023 at 04:59:08PM +0200, daggs wrote: >> >> >> >Greetings Martin, >> >> >> > >> >> >> >> Sent: Sunday, October 22, 2023 at 12:37 PM >> >> >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> >> >> >> To: "daggs" <daggs(a)gmx.com&gt; >> >> >> >> Cc: libvir-list(a)redhat.com >> >> >> >> Subject: Re: hdd kills vm >> >> >> >> >> >> >> >> On Fri, Oct 20, 2023 at 02:42:38PM +0200, daggs wrote: >> >> >> >> >Greetings, >> >> >> >> > >> >> >> >> >I have a windows 11 vm running on my Gentoo using libvirt (9.8.0) + qemu (8.1.2), I'm passing almost all available resources to the vm >> >> >> >> >(all 16 cpus, 31 out of 32 GB, nVidia gpu is pt), but the performance is not good, system lags, takes long time to boot. >> >> >> >> >> >> >> >> There are couple of things that stand out to me in your setup and I'll >> >> >> >> assume the host has one NUMA node with 8 cores, each with 2 threads as, >> >> >> >> just like you set it up in the guest XML. >> >> >> >thats correct, see: >> >> >> >$ lscpu | grep -i numa >> >> >> >NUMA node(s): 1 >> >> >> >NUMA node0 CPU(s): 0-15 >> >> >> > >> >> >> >however: >> >> >> >$ dmesg | grep -i numa >> >> >> >[ 0.003783] No NUMA configuration found >> >> >> > >> >> >> >can that be the reason? >> >> >> > >> >> >> >> >> >> no, this is fine, 1 NUMA node is not a NUMA, technically, so that's >> >> >> perfectly fine. >> >> >thanks for clarifying it for me >> >> > >> >> >> >> >> >> >> >> >> >> >> * When you give the guest all the CPUs the host has there is nothing >> >> >> >> left to run the host tasks. You might think that there "isn't >> >> >> >> anything running", but there is, if only your init system, the kernel >> >> >> >> and the QEMU which is emulating the guest. This is definitely one of >> >> >> >> the bottlenecks. >> >> >> >I've tried with 12 out of 16, same behavior. >> >> >> > >> >> >> >> >> >> >> >> * The pinning of vCPUs to CPUs is half-suspicious. If you are trying to >> >> >> >> make vCPU 0 and 1 be threads on the same core and on the host the >> >> >> >> threads are represented as CPUs 0 and 8, then that's fine. If that is >> >> >> >> just copy-pasted from somewhere, then it might not reflect the current >> >> >> >> situation and can be source of many scheduling issues (even once the >> >> >> >> above is dealt with). >> >> >> >I found a site that does it for you, if it is wrong, can you point me to a place I can read about it? >> >> >> > >> >> >> >> >> >> Just check what the topology is on the host and try to match it with the >> >> >> guest one. If in doubt, then try it without the pinning. >> >> >I can try to play with it, what I don't know is what should be the mapping logic? >> >> > >> >> >> >> Threads on the same core in the guest should map to threads on the same >> >> core in the host. Since there is no NUMA that should be enough to get >> >> the best performance. But even misconfiguration of this will not >> >> introduce lags in the system if it has 8 CPUs. So that's definitely not >> >> the root cause of the main problem, it just might be suboptimal. >> >> >> >> >> >> >> >> >> >> >> >> >> * I also seem to recall that Windows had some issues with systems that >> >> >> >> have too many cores. I'm not sure whether that was an issue with an >> >> >> >> edition difference or just with some older versions, or if it just did >> >> >> >> not show up in the task manager, but there was something that was >> >> >> >> fixed by using either more sockets or cores in the topology. This is >> >> >> >> probably not the issue for you though. >> >> >> >> >> >> >> >> >after trying a few ways to fix it, I've concluded that the issue might be related to the why the hdd is defined at the vm level. >> >> >> >> >here is the xml: https://bpa.st/MYTA >> >> >> >> >I assume that the hdd sits on the sata ctrl causing the issue but I'm not sure what is the proper way to fix it, any ideas? >> >> >> >> > >> >> >> >> >> >> >> >> It looks like your disk is on SATA, but I don't see why that would be an >> >> >> >> issue. Passing the block device to QEMU as VirtIO shouldn't cause that >> >> >> >> much of a difference. Try measuring the speed of the disk on the host >> >> >> >> and then in the VM maybe. Is that SSD or NVMe? I presume that's not >> >> >> >> spinning rust, is it. >> >> >> >as seen, I have 3 drives, 2 cdroms as sata and one hdd pt as virtio, I read somewhere that if the controller of the virtio >> >> >> >device is sata, than it doesn't uses the virtio optimally. >> >> >> >> >> >> Well it _might_ be slightly more beneficial to use virtio-scsi or even >> >> >> <disk type='block' device='lun'>, but I can't imagine that would make >> >> >> the system lag. I'm not that familiar with the details. >> >> >configure virtio-scsi and sata-scai at the same time? >> >> > >> >> >> >> Yes, forgot that, sorry. Try virtio-scsi. You could also go farther >> >> and pass through the LUN or the whole HBA (if you don't need to access >> >> any other disk on it) to the VM. Try the information presented here: >> >> >> >> https://libvirt.org/formatdomain.html#usb-pci-scsi-devices >> >> >> >> >> >> >> >> >it is a spindle, nvmes are too expensive where I live, frankly, I don't need lightning fast boot, the other BM machines running windows on spindle >> >> >> >run it quite fast and they aren't half as fast as this server >> >> >> > >> >> >> >> >> >> That might actually be related. The guest might think it is a different >> >> >> type of disk and use completely suboptimal scheduling. This might >> >> >> actually be solved by passing it as <disk device='lun'..., but at this >> >> >> point I'm just guessing. >> >> >I'll look into that, thanks. >> > >> >so bottom line, you suggest the following: >> >1. remove the manual cpu pin, let qemu sort that out. >> >> You might try it, of course pinning it is in the end the better option. >> >> >2. add a virtio scsi controller and connect the os hdd to it >> >3. pass the hss via scsi pt and not dev node >> >4. if I able to do #3, no need to add device='lun' as it won't use the disk option >> > >> >> First try (3), then you don't need to do anything else and if that >> succeeds you have the superior configuration. If you can pass through >> something that will not remove anything from your host system. >> >> >Dagg. >> > >> > >I've decided to first try #3 as yo suggested, based on this output: >$ lsscsi >[0:0:0:0] disk ATA WDC WD1003FZEX-0 1A01 /dev/sda >[1:0:0:0] disk ATA WDC WD10EZEX-08W 1A02 /dev/sdb >[2:0:0:0] disk ATA SAMSUNG HD103SJ 0001 /dev/sdc >[3:0:0:0] disk ATA SAMSUNG HD103SJ 0001 /dev/sdd >[4:0:0:0] disk ATA ST1000DM005 HD10 00E5 /dev/sde >[5:0:0:0] disk ATA WDC WD10EZEX-08W 1A02 /dev/sdf >[6:0:0:0] disk Kingston DataTraveler 3.0 0000 /dev/sdg >[7:0:0:0] cd/dvd TS8XDVDS TRANSCEND 1.02 /dev/sr0 > >I deduced my data is 0:0:0:0, so I've added this to the file: I have to trust you here, the link to the XML does not lead anywhere at the moment.

><controller type='scsi' index='0' model='virtio-scsi'> > <address type='pci' domain='0x0000' bus='0x00' slot='0x0c' function='0x0'/> ></controller> ><hostdev mode='subsystem' type='scsi' managed='no'> With managed='no' you are responsible for detaching and re-attaching the device for it to be accessible to QEMU. With managed='yes' libvirt can do that for you. But be really really sure that it is the device you want to plug to the guest domain.

Sent: Friday, November 03, 2023 at 1:15 PM From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; To: "daggs" <daggs(a)gmx.com&gt; Cc: users(a)lists.libvirt.org Subject: Re: hdd kills vm On Thu, Nov 02, 2023 at 06:06:24PM +0100, daggs wrote: > > >> Sent: Thursday, November 02, 2023 at 2:34 PM >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> To: "daggs" <daggs(a)gmx.com&gt; >> Cc: users(a)lists.libvirt.org >> Subject: Re: hdd kills vm >> >> On Wed, Nov 01, 2023 at 10:24:05PM +0100, daggs wrote: >> >> Sent: Wednesday, November 01, 2023 at 10:06 AM >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; >> >> To: "daggs" <daggs(a)gmx.com&gt; >> >> Cc: users(a)lists.libvirt.org >> >> Subject: Re: hdd kills vm >> >> [...] >> >> >so bottom line, you suggest the following: >> >> >1. remove the manual cpu pin, let qemu sort that out. >> >> >> >> You might try it, of course pinning it is in the end the better option. >> >> >> >> >2. add a virtio scsi controller and connect the os hdd to it >> >> >3. pass the hss via scsi pt and not dev node >> >> >4. if I able to do #3, no need to add device='lun' as it won't use the disk option >> >> > >> >> >> >> First try (3), then you don't need to do anything else and if that >> >> succeeds you have the superior configuration. If you can pass through >> >> something that will not remove anything from your host system. >> >> >> >> >Dagg. >> >> > >> >> >> > >> >I've decided to first try #3 as yo suggested, based on this output: >> >$ lsscsi >> >[0:0:0:0] disk ATA WDC WD1003FZEX-0 1A01 /dev/sda >> >[1:0:0:0] disk ATA WDC WD10EZEX-08W 1A02 /dev/sdb >> >[2:0:0:0] disk ATA SAMSUNG HD103SJ 0001 /dev/sdc >> >[3:0:0:0] disk ATA SAMSUNG HD103SJ 0001 /dev/sdd >> >[4:0:0:0] disk ATA ST1000DM005 HD10 00E5 /dev/sde >> >[5:0:0:0] disk ATA WDC WD10EZEX-08W 1A02 /dev/sdf >> >[6:0:0:0] disk Kingston DataTraveler 3.0 0000 /dev/sdg >> >[7:0:0:0] cd/dvd TS8XDVDS TRANSCEND 1.02 /dev/sr0 >> > >> >I deduced my data is 0:0:0:0, so I've added this to the file: >> >> I have to trust you here, the link to the XML does not lead anywhere at the moment. >> >used this tutorial: https://github.com/stevenewbs/blog/blob/master/libvirtd%20qemu%20Blu%20Ra... > >new link is here: https://bpa.st/SU4Q > >> ><controller type='scsi' index='0' model='virtio-scsi'> >> > <address type='pci' domain='0x0000' bus='0x00' slot='0x0c' function='0x0'/> >> ></controller> >> ><hostdev mode='subsystem' type='scsi' managed='no'> >> >> With managed='no' you are responsible for detaching and re-attaching the device >> for it to be accessible to QEMU. With managed='yes' libvirt can do that for >> you. But be really really sure that it is the device you want to plug to the >> guest domain. >so it should work if I set managed to 'yes'? > I think so, make sure it is the right disk. You could also pass through the whole HBA with type='scsi_host' since it looks like there is nothing else on that one IIUC.

Sent: Friday, November 03, 2023 at 2:43 PM From: "daggs via Users" <users(a)lists.libvirt.org&gt; To: "Martin Kletzander" <mkletzan(a)redhat.com&gt; Cc: users(a)lists.libvirt.org Subject: Re: hdd kills vm > Sent: Friday, November 03, 2023 at 2:04 PM > From: "daggs via Users" <users(a)lists.libvirt.org&gt; > To: "Martin Kletzander" <mkletzan(a)redhat.com&gt; > Cc: users(a)lists.libvirt.org > Subject: Re: hdd kills vm > > > > > Sent: Friday, November 03, 2023 at 1:15 PM > > From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; > > To: "daggs" <daggs(a)gmx.com&gt; > > Cc: users(a)lists.libvirt.org > > Subject: Re: hdd kills vm > > > > On Thu, Nov 02, 2023 at 06:06:24PM +0100, daggs wrote: > > > > > > > > >> Sent: Thursday, November 02, 2023 at 2:34 PM > > >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; > > >> To: "daggs" <daggs(a)gmx.com&gt; > > >> Cc: users(a)lists.libvirt.org > > >> Subject: Re: hdd kills vm > > >> > > >> On Wed, Nov 01, 2023 at 10:24:05PM +0100, daggs wrote: > > >> >> Sent: Wednesday, November 01, 2023 at 10:06 AM > > >> >> From: "Martin Kletzander" <mkletzan(a)redhat.com&gt; > > >> >> To: "daggs" <daggs(a)gmx.com&gt; > > >> >> Cc: users(a)lists.libvirt.org > > >> >> Subject: Re: hdd kills vm > > >> >> > > [...] > > >> >> >so bottom line, you suggest the following: > > >> >> >1. remove the manual cpu pin, let qemu sort that out. > > >> >> > > >> >> You might try it, of course pinning it is in the end the better option. > > >> >> > > >> >> >2. add a virtio scsi controller and connect the os hdd to it > > >> >> >3. pass the hss via scsi pt and not dev node > > >> >> >4. if I able to do #3, no need to add device='lun' as it won't use the disk option > > >> >> > > > >> >> > > >> >> First try (3), then you don't need to do anything else and if that > > >> >> succeeds you have the superior configuration. If you can pass through > > >> >> something that will not remove anything from your host system. > > >> >> > > >> >> >Dagg. > > >> >> > > > >> >> > > >> > > > >> >I've decided to first try #3 as yo suggested, based on this output: > > >> >$ lsscsi > > >> >[0:0:0:0] disk ATA WDC WD1003FZEX-0 1A01 /dev/sda > > >> >[1:0:0:0] disk ATA WDC WD10EZEX-08W 1A02 /dev/sdb > > >> >[2:0:0:0] disk ATA SAMSUNG HD103SJ 0001 /dev/sdc > > >> >[3:0:0:0] disk ATA SAMSUNG HD103SJ 0001 /dev/sdd > > >> >[4:0:0:0] disk ATA ST1000DM005 HD10 00E5 /dev/sde > > >> >[5:0:0:0] disk ATA WDC WD10EZEX-08W 1A02 /dev/sdf > > >> >[6:0:0:0] disk Kingston DataTraveler 3.0 0000 /dev/sdg > > >> >[7:0:0:0] cd/dvd TS8XDVDS TRANSCEND 1.02 /dev/sr0 > > >> > > > >> >I deduced my data is 0:0:0:0, so I've added this to the file: > > >> > > >> I have to trust you here, the link to the XML does not lead anywhere at the moment. > > >> > > >used this tutorial: https://github.com/stevenewbs/blog/blob/master/libvirtd%20qemu%20Blu%20Ra... > > > > > >new link is here: https://bpa.st/SU4Q > > > > > >> ><controller type='scsi' index='0' model='virtio-scsi'> > > >> > <address type='pci' domain='0x0000' bus='0x00' slot='0x0c' function='0x0'/> > > >> ></controller> > > >> ><hostdev mode='subsystem' type='scsi' managed='no'> > > >> > > >> With managed='no' you are responsible for detaching and re-attaching the device > > >> for it to be accessible to QEMU. With managed='yes' libvirt can do that for > > >> you. But be really really sure that it is the device you want to plug to the > > >> guest domain. > > >so it should work if I set managed to 'yes'? > > > > > > > I think so, make sure it is the right disk. You could also pass through the > > whole HBA with type='scsi_host' since it looks like there is nothing else on > > that one IIUC. > I'll try with managed=yes tried that, didn't booted. any idea what am I doing wrong?