Hi, I think that one of the best things from Proxmox is the ability to connect to any of the hosts via web and been able to admin the hole infrastructure from there. This avoids the single point of failure that today engine represents in ovirt plataform.

I agree with Jakub that been able to mix Intel and AMD hosts would be great so we can use the hole of the servers into a DC and engine may migrate VM to the same kind of processors if available but if not,

migrate to any available host and alert about the performance degradation.

I'm still a rookie to ovirt so at this time I don't know all the features to make further coments. I think you have a great product and the best of it is the speed it's improving.

Thanks for the info. Regards, On 06/10/13 17:26, Itamar Heim wrote: > On 08/29/2013 05:22 PM, Juan Pablo Lorier wrote: >> Hi, >> >> I think that one of the best things from Proxmox is the ability to >> connect to any of the hosts via web and been able to admin the hole >> infrastructure from there. This avoids the single point of failure that >> today engine represents in ovirt plataform. > hosted engine should resolve this one. > >> I agree with Jakub that been able to mix Intel and AMD hosts would be >> great so we can use the hole of the servers into a DC and engine may >> migrate VM to the same kind of processors if available but if not, > as i replied just now, afaik, live migration between intel/amd isn't > supported by kvm. > >> migrate to any available host and alert about the performance >> degradation. > you can't really migrate to a "lesser" host, only to at least same cpu > level. > >> I'm still a rookie to ovirt so at this time I don't know all the >> features to make further coments. I think you have a great product and >> the best of it is the speed it's improving. > thanks, glad to hear. > _______________________________________________ Users mailing list Users(a)ovirt.org http://lists.ovirt.org/mailman/listinfo/users

On 11/06/2013 06:41 PM, Jakub Bittner wrote: > Dne 7.10.2013 13:58, Juan Pablo Lorier napsal(a): >> Thanks for the info. >> Regards, >> >> On 06/10/13 17:26, Itamar Heim wrote: >>> On 08/29/2013 05:22 PM, Juan Pablo Lorier wrote: >>>> Hi, >>>> >>>> I think that one of the best things from Proxmox is the ability to >>>> connect to any of the hosts via web and been able to admin the hole >>>> infrastructure from there. This avoids the single point of failure >>>> that >>>> today engine represents in ovirt plataform. >>> hosted engine should resolve this one. >>> >>>> I agree with Jakub that been able to mix Intel and AMD hosts would be >>>> great so we can use the hole of the servers into a DC and engine may >>>> migrate VM to the same kind of processors if available but if not, >>> as i replied just now, afaik, live migration between intel/amd isn't >>> supported by kvm. >>> >>>> migrate to any available host and alert about the performance >>>> degradation. >>> you can't really migrate to a "lesser" host, only to at least same cpu >>> level. >>> >>>> I'm still a rookie to ovirt so at this time I don't know all the >>>> features to make further coments. I think you have a great product >>>> and >>>> the best of it is the speed it's improving. >>> thanks, glad to hear. >>> >> _______________________________________________ >> Users mailing list >> Users(a)ovirt.org >> http://lists.ovirt.org/mailman/listinfo/users > Hello, > > sorry for reopening old topic, but I have 1 (or 2) feature request to > ovirt 3.4. It would be great to have possibility to edit network > description and comment while it is attached to running VM. Second please open an BZ RFE, maybe add to the 3.4 planning doc? > feature request is fault tolerant VM. I did not find any info about this > in ovirt discussion. Basically it means running same VM on two(or more) > separated host where one is standard VM and the second one is exact copy > (network, ram contents, disk, everything). When host where standard VM > runs fails, then the mirrored VM should became standard one and continue > operation without any user notice. Like HA, but without restart. I think we're still waiting on this to be more robust in qemu-kvm first > > Thank you. > _______________________________________________ > Users mailing list > Users(a)ovirt.org > http://lists.ovirt.org/mailman/listinfo/users