|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Xen-users] Re: how to congure dom0 to 1G, and domU to 1G
On Sat, May 22, 2010 at 12:06:37AM +0800, lei yang wrote: > On Sat, May 22, 2010 at 12:01 AM, Pasi Kärkkäinen <[1]pasik@xxxxxx> wrote: > > On Fri, May 21, 2010 at 11:43:48PM +0800, lei yang wrote: > > On Fri, May 21, 2010 at 8:08 PM, Pasi Kärkkäinen > <[1][2]pasik@xxxxxx> wrote: > > > > On Thu, May 20, 2010 at 04:59:00PM +0800, lei yang wrote: > > > On Thu, May 20, 2010 at 2:04 PM, Pasi Kärkkäinen > > <[1][2][3]pasik@xxxxxx> wrote: > > > > > > On Thu, May 20, 2010 at 01:59:16PM +0800, lei yang wrote: > > > > Hi Experts, > > > > > > > > > > I removed xen-devel from CC since this is a very basic > question, > > > and not related to Xen development. > > > > I want to configure dom0 to 1g and dom0 to 1g ? > > > > > > > > I know dom0_mem=1024M can do it for dom0, how about > domU? > > > > > > > > > > Uhm.. for domU you specify: > > > > > > memory = 1024 > > > > > > in the domU configuration file (/etc/xen/<guest>). > > > > My machine has 8 cores > > > > how to specify xen to 4 cpus? > > > > > > > > > > Xen is the *hypervisor*, so it has to use ALL the physical > > CPUs/cores, > > > to be able to schedule vcpus!! > > > > > > Do you mean we can't specify xen to 4 cpus, and if we > specify > > "nosmp" what > > > does it mean > > > it will use *just* one physical core to schedule vcpus? and > dom0 > > and domU > > > just can use one core? > > > > > > > Depends where you specify "nosmp", for Xen or for dom0 kernel. > > > > > > if I specify to nosmp to "xen", will dom0 and domU just run core > one core? > > > > I think so. Why would you want to limit everything to just one core ? > > > > I want to test the xen UP mode performance. so I need to run everything > just on one core. if I specify xen to "nosmp" > dom0 and domU will just run on one core, how about xen? just run on one > core? > I assume everything runs on a single core then. You can verify with "xm info". It shows what Xen hypervisor has/uses. -- Pasi > Thanks > Lei > > > > > can we now which core? > > > > I don't think so. > > -- Pasi > > > > Xen *has* to have access to all the physical cpus/cores you want > to use > > for your guest, including dom0. > > > > See Xen vcpu pinning for more information. > > > > > > > > > > how to specify dom0 4 cpus > > > > > > > > > > dom0_max_vcpus=4 for xen.gz in grub.conf. > > > > > > > > > > > > Specified options "dom0_max_vcpus=2" and "dom0_vcpus_pin". > > > Specified option "dom0_max_vcpus=2". > > > > > > what's the difference? which performance will be higher? > > > > > > > Adding dom0_vcpus_pin will automatically pin the dom0 vcpus to > matching > > physical cpus. > > That might not make a difference alone.. you need/want to pin the > other > > guest vcpus aswell, > > depending what kind of configuration you want to get. > > > > > > "dom0_max_vcpus=2" and "dom0_vcpus_pin". core0 and core 1 will > just used > > by dom0? can't be used by others eg: domU? > > > > -- Pasi > > > > > > > > Thanks > > > Lei > > > > > > > how to specify domU 5 cpus ? > > > > > > > > > > vcpus = 5 > > > > can we specify dom0 to 1 cpu, at same time domU to 6 > cpu > > > > > > > > > > dom0_max_vcpus=1 dom0_vcpus_pin > > > > > > in grub.conf for xen.gz. > > > > can we specify nosmp to xen, and then specify dom0 > cpu > 1? > > > > > > > > > > No. > > > > > > See: > [2][3][4]http://wiki.xensource.com/xenwiki/XenBestPractices > > > and: > [3][4][5]http://wiki.xensource.com/xenwiki/XenCommonProblems > > > -- Pasi > > > > > > > > > > > > Thanks > > > > Lei > > > > > > > > -- > > > > "We learn from failure, not from success!" > > > > > > -- > > > "We learn from failure, not from success!" > > > > > > References > > > > > > Visible links > > > 1. mailto:[5][6]pasik@xxxxxx > > > 2. [6][7]http://wiki.xensource.com/xenwiki/XenBestPractices > > > 3. [7][8]http://wiki.xensource.com/xenwiki/XenCommonProblems > > > > -- > > "We learn from failure, not from success!" > > > > References > > > > Visible links > > 1. mailto:[9]pasik@xxxxxx > > 2. mailto:[10]pasik@xxxxxx > > 3. [11]http://wiki.xensource.com/xenwiki/XenBestPractices > > 4. [12]http://wiki.xensource.com/xenwiki/XenCommonProblems > > 5. mailto:[13]pasik@xxxxxx > > 6. [14]http://wiki.xensource.com/xenwiki/XenBestPractices > > 7. [15]http://wiki.xensource.com/xenwiki/XenCommonProblems > > -- > "We learn from failure, not from success!" > > References > > Visible links > 1. mailto:pasik@xxxxxx > 2. mailto:pasik@xxxxxx > 3. mailto:pasik@xxxxxx > 4. http://wiki.xensource.com/xenwiki/XenBestPractices > 5. http://wiki.xensource.com/xenwiki/XenCommonProblems > 6. mailto:pasik@xxxxxx > 7. http://wiki.xensource.com/xenwiki/XenBestPractices > 8. http://wiki.xensource.com/xenwiki/XenCommonProblems > 9. mailto:pasik@xxxxxx > 10. mailto:pasik@xxxxxx > 11. http://wiki.xensource.com/xenwiki/XenBestPractices > 12. http://wiki.xensource.com/xenwiki/XenCommonProblems > 13. mailto:pasik@xxxxxx > 14. http://wiki.xensource.com/xenwiki/XenBestPractices > 15. http://wiki.xensource.com/xenwiki/XenCommonProblems _______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-users
|
 |
Lists.xenproject.org is hosted with RackSpace, monitoring our |