similar to: How to retrieve legacy cgroups location ?

Hi, My piece of code ( C langage) uses cgroups to retrieve counters related to cpu and memory usage related to KVM guests hosted by the host where this code runs. I noticed that depending on the O.S. running on the host , these counters are not found at the same location : CentOS 7 : ls /sys/fs/cgroup/cpuacct/machine.slice/machine-qemu\\x2drhel6.0.scope/vcpu0 cgroup.clone_children

On Thu, Oct 29, 2015 at 10:40:44AM +0000, Jean-Pierre Ribeauville wrote: >Hi, > >As told in "Control Groups Resource Management" libvirt page : >Legacy cgroups layout >Prior to libvirt 1.0.5, the cgroups layout created by libvirt was different from that described above, and did not allow for administrator customization. Libvirt used a fixed, 3-level hierarchy

Hi, 1) is it enough to add a hook.py in /usr/libexec/vdsm/hooks/before_vm_start directory , and then shut down and reboot a guest to see this hook.py invoked ? 2) when running manually my hook.py, I got following error : ImportError: No module named hooking Do I have to install anything to solve this issue ? Thanks for help. Regards, J.P. Ribeauville P:

Hi, I'm trying to find out what is the minimum release I need to use to have this field 'state' available ? <channel type='unix'> <source mode='bind' path='//var/lib/libvirt/qemu/dummy_agent'/> <target type='virtio' name=dummy' state='connected'/> <alias name='channel0'/> <address

Hi , Despite the fact that my Guest has an IP address , by running this piece of code on a KVM host : ifaces_count = F_virDomainInterfaceAddresses(domain, &ifaces,VIR_DOMAIN_INTERFACE_ADDRESSES_SRC_LEASE, NULL); ifaces_count = F_virDomainInterfaceAddresses(domain, &ifaces,VIR_DOMAIN_INTERFACE_ADDRESSES_SRC_AGENT, NULL); I'm not able to get a ifaces_count different from

Hi, I' trying to read datas sent by my KVM host within the Guest. Host send these datas trough a channel UNIX type , by using AF_UNIX socket stuff. If i try to retrive the datas by using such type of thing in the Guest : #define NAME "/dev/virtio-ports/agent" sock = socket(AF_UNIX, SOCK_STREAM, 0); 12 if (sock < 0) { 13 perror("opening stream socket");

Hi, By adding following lines in a Guest .xml conf file , I'm able to send / receive datas between host and this guest. <channel type='pty'> <target type='virtio' name='arbitrary.virtio.serial.port.name'/> <address type='virtio-serial' controller='0' bus='0' port='1'/> </channel> My question is

>5. LTP testing might run very slow in SMP VTI Domain with credit scheduler. If binding VTI >and Xen0 vcpu, this bug won''t be there. Hi keir, In credit scheduler, two vcpus in the same domain may be scheduled on the same CPU. For instance, vcpu0 and vcpu1 are running on the same CPU, vcpu0 is doing spin_lock in guest, then time slice is due, vcpu0 is scheduled out before doing

Hi, I'm trying to use unix-type channel to send/receive datas between host and guest . Here are the steps I follow : 1) Adding these lines to expose a channel unix type to my guest <channel type='unix'> <source mode='bind' path='/var/lib/libvirt/qemu/omnivision_1.agent'/> <target type='virtio'

Hi, I found an answer : Use virConnectOpenAuth() instead of virConnectOpenReadOnly(). As my piece of code is running silently (without user connected) , by using virConnectOpenAuth() , is it possible to avoid to prompt user to get user /password ? Thx. J.P. -----Message d'origine----- De : libvirt-users-bounces@redhat.com [mailto:libvirt-users-bounces@redhat.com] De la part de

Hi, In which libvirt version are these virDomainMemoryStats() counters handled ? VIR_DOMAIN_MEMORY_STAT_SWAP_OUT VIR_DOMAIN_MEMORY_STAT_MAJOR_FAULT VIR_DOMAIN_MEMORY_STAT_MINOR_FAULT VIR_DOMAIN_MEMORY_STAT_UNUSED VIR_DOMAIN_MEMORY_STAT_AVAILABLE Thanks. Regards, J.P. Ribeauville P: +33.(0).1.47.17.20.49 . Puteaux 3 Etage 5 Bureau 4

Hi, Is there an equivalent of "virsh attach-device -live -persistent " within libvirt API : virDomainAttachDevice() or virDomainAttachDeviceFlags() or ??? Thanks for help. Regards, J.P. Ribeauville P: +33.(0).1.47.17.20.49 . Puteaux 3 Etage 5 Bureau 4 jpribeauville@axway.com<mailto:jpribeauville@axway.com> http://www.axway.com<http://www.axway.com/> P

Hi, I think that issue is due to the fact that my Guest is a transient domain. When shutting down it or migrating to another host , then virsh list --all doesn't show this Guest anymore. How may I make this Guest a persistent one ? ( i.e. this Guest has been created via RHEV-M GUI) Thx for hlep. Regards, J.P. -----Message d'origine----- De : libvirt-users-bounces@redhat.com

Currently, the getdomaininfo function (used to fill in a dom0_getdomaininfo_t for a domain) calculates a domain''s total CPU time from its VCPU times using the code: if ( v->cpu_time > cpu_time ) cpu_time += v->cpu_time; This causes a VCPU''s time to only be counted if it is greater than the current total; so if VCPU0 has 10 seconds and VCPU1 has 5, the

Hi, When using « virsh attach-device ....." to add a channel device to a Guest , I see following entry in libvirt log : 2015-12-07 11:14:47.307+0000: 14551: debug : virDomainAttachDeviceFlags:10433 : dom=0x7f2a58252fc0, ........ Great !! But , when calling virDomainAttachDeviceFlags( ) from a C program , I got return code -1 but nothing appears related to virDomainAttachDeviceFlags(

Hi, I need to add , via script , a channel device in .xml guest config file. Is virsh edit useable from a "standard" script or is there any more suitable way to use ? Thanks. J.P. Ribeauville P: +33.(0).1.47.17.20.49 . Puteaux 3 Etage 5 Bureau 4 jpribeauville at axway.com<mailto:jpribeauville at axway.com> http://www.axway.com<http://www.axway.com/> P Pensez ?

I''ve been looking into bug [1]#195 and I have a couple of questions on how timer interrupts and virqs are handled. Is it possible for dom0 linux to irq balance timer interrupts to different cpus? That is, if xen sends a VIRQ_TIMER to vcpu0, backed by cpu0, is it possible for that interrupt to be handled by vcpu1, backed by cpu1 ? After putting in some debug code in to timer_interrupt

Hi, Is a change , by edit it via virsh edit , in the "local" .xml configuration file (i.e. adding a channel device) persistent when live migrating the guest on another host ? Thx. J.P. Ribeauville P: +33.(0).1.47.17.27.87 Puteaux 3 Etage 5 Bureau 4 jpribeauville@axway.com<mailto:jpribeauville@axway.com> http://www.axway.com<http://www.axway.com/> P Pensez à

Hello, Does it exist some documentation concerning following topic ? What's difference between following perf counters : /cgroup/cpuacct/libvirt/qemu/test/cpuacct.usage and /cgroup/cpuacct/libvirt/qemu/test/emulator/cpuacct.usage Thx . Regards, J.P. Ribeauville P: +33.(0).1.47.17.27.78 Puteaux 3 Etage 5 Bureau 4 jpribeauville@axway.com<mailto:jpribeauville@axway.com>

Hi, As I want to start qemu by adding following parameters : -device virtio-serial -chardev socket,path=/tmp/foo,server,nowait,id=test -device virtserialport,chardev=test,name=port.0 Is it possible to add them in the .xml configuration file of the guest , so they will be automatically added when the guest is started ? If yes , any doc describing the syntax ? Thx. Regards, J.P. Ribeauville