[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
if you are eligible for the developr subscription you can subscribe at : https://developers.redhat.com/register P.S.: It now includes up to 16 RHEL machines for production usage. Best Regards,Strahil Nikolov On Sun, Oct 3, 2021 at 12:57, Tommy Sway wrote: ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/7O3MGF5IGJQJ4SP7ILZU4NI2JAVGMK4Y/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/RIR2DJOGDMNXAXDDNYQZVEGSI7LWKN6I/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
SUBSCRIBER EXCLUSIVE CONTENT An active Red Hat subscription is required to participate. I cannot get the result. From: users-boun...@ovirt.org On Behalf Of Strahil Nikolov via Users Sent: Sunday, October 3, 2021 4:03 AM To: sz_cui...@163.com; 'Strahil Nikolov via Users' Subject: [ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit vdsm supports dynamic allocation of Huge Pages, but https://access.redhat.com/solutions/4904441 indicates the issues in 4.3 related to hugepages (you can see the solution via RH dev subscription). If you needed to set them manually and everything works normally, I would have picked something like 90-95% in HugePages. Yet, due to the issues described in that solution - it's better to enable HugePages only in the Guest. Best Regards, Strahil Nikolov On Sat, Oct 2, 2021 at 10:13, Tommy Sway mailto:sz_cui...@163.com> > wrote: ___ Users mailing list -- users@ovirt.org <mailto:users@ovirt.org> To unsubscribe send an email to users-le...@ovirt.org <mailto:users-le...@ovirt.org> Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/CEXZE5QYTWWNMFQU5ASJR75IWOOIFYLJ/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/7O3MGF5IGJQJ4SP7ILZU4NI2JAVGMK4Y/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
For 4.3 - yes. This is due to the issues in 4.3Once you move to 4.4 , you should be able to enable HugePages (not THP) on the Hypervisors without troubles. I assume that your workload is 'in production' and stability (no downtime) is more important that some performance gain from HugePages on the Hypervisors. @Simon, any hints when OLVM 4.4 will be released ? Best Regards,Strahil Nikolov On Sun, Oct 3, 2021 at 6:29, tommy sway wrote: ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/KLHJDZMPLKPPXABYSP55DA472UPCKCDO/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/4ULRPOJFX2SI4ZWQJYGZ2NA6FO7GTOIN/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
Do you mean the best way is to enable traditional hugepage memory only on virtual machines and turn off the default transparent hugepage memory on physical machines and without enabling traditional hugepage memory? On 10/03/2021 04:03, Strahil Nikolov via Users wrote: vdsm supports dynamic allocation of Huge Pages, but https://access.redhat.com/solutions/4904441 indicates the issues in 4.3 related to hugepages (you can see the solution via RH dev subscription). If you needed to set them manually and everything works normally, I would have picked something like 90-95% in HugePages. Yet, due to the issues described in that solution - it's better to enable HugePages only in the Guest. Best Regards, Strahil Nikolov On Sat, Oct 2, 2021 at 10:13, Tommy Sway wrote: ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/CEXZE5QYTWWNMFQU5ASJR75IWOOIFYLJ/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/KLHJDZMPLKPPXABYSP55DA472UPCKCDO/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
vdsm supports dynamic allocation of Huge Pages, but https://access.redhat.com/solutions/4904441 indicates the issues in 4.3 related to hugepages (you can see the solution via RH dev subscription).If you needed to set them manually and everything works normally, I would have picked something like 90-95% in HugePages. Yet, due to the issues described in that solution - it's better to enable HugePages only in the Guest. Best Regards,Strahil Nikolov On Sat, Oct 2, 2021 at 10:13, Tommy Sway wrote: ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/CEXZE5QYTWWNMFQU5ASJR75IWOOIFYLJ/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/IMERYEBAAAFH5GRXTVLF6FDLPTWBWNXQ/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
If it is transparent huge page, I can only enable it, but the traditional huge page is required to specify the specific size. Should I follow the principle that the size of huge pages on a physical machine equals the sum of the size of huge pages on all VMS running on it? From: users-boun...@ovirt.org On Behalf Of tommy sway Sent: Saturday, October 2, 2021 1:32 PM To: Strahil Nikolov via Users Subject: [ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit but config what size of the hugepage on the kvm host? On 10/02/2021 01:33, <mailto:users@ovirt.org> Strahil Nikolov via Users wrote: KVM host -> Just enable it. All VMs will be in Hugepages. VM -> just for the SGA. Best Regards, Strahil Nikolov On Fri, Oct 1, 2021 at 18:01, tommy mailto:sz_cui...@163.com> > wrote: If both the KVM server and the VIRTUAL machine are configured with traditional large pages, there is also the issue of configuring size. Since only THE SGA uses large pages, other database components cannot use them. Should the KVM server and virtual machine use the same memory ratio for large pages? For example, use 50% of the total memory size to avoid imbalance caused by improper proportion. 原始邮件 发件人: Strahil Nikolov mailto:hunter86...@yahoo.com> > 日期: 2021年10月1日周五 22:52 收件人: tommy sway mailto:sz_cui...@163.com> >, Strahil Nikolov via Users mailto:users@ovirt.org> > 主 题: Re: [ovirt-users]Re: 回复: Re: About the vm memory limit Actually yes - should speedup the process. I've edited that draft too many times :) On Fri, Oct 1, 2021 at 7:32, tommy sway mailto:sz_cui...@163.com> > wrote: ___ Users mailing list -- users@ovirt.org <mailto:users@ovirt.org> To unsubscribe send an email to users-le...@ovirt.org <mailto:users-le...@ovirt.org> Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/LMPJISCMPKRWJN4KW5DVVWEXLRK45L7L/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/CEXZE5QYTWWNMFQU5ASJR75IWOOIFYLJ/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
but config what size of the hugepage on the kvm host? On 10/02/2021 01:33, Strahil Nikolov via Users wrote: KVM host -> Just enable it. All VMs will be in Hugepages. VM -> just for the SGA. Best Regards, Strahil Nikolov On Fri, Oct 1, 2021 at 18:01, tommy wrote: If both the KVM server and the VIRTUAL machine are configured with traditional large pages, there is also the issue of configuring size. Since only THE SGA uses large pages, other database components cannot use them. Should the KVM server and virtual machine use the same memory ratio for large pages? For example, use 50% of the total memory size to avoid imbalance caused by improper proportion. 原始邮件 发件人: Strahil Nikolov 日期: 2021年10月1日周五 22:52 收件人: tommy sway , Strahil Nikolov via Users 主 题: Re: [ovirt-users]Re: 回复: Re: About the vm memory limit Actually yes - should speedup the process. I've edited that draft too many times :) On Fri, Oct 1, 2021 at 7:32, tommy sway wrote: ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/LMPJISCMPKRWJN4KW5DVVWEXLRK45L7L/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/JHTZU36VPEC53JNOC6AAKH7INQIY3OHH/
[ovirt-users]Re: 回复:Re: 回复: Re: About the vm memory limit
KVM host -> Just enable it. All VMs will be in Hugepages. VM -> just for the SGA. Best Regards,Strahil Nikolov On Fri, Oct 1, 2021 at 18:01, tommy wrote: If both the KVM server and the VIRTUAL machine are configured with traditional large pages, there is also the issue of configuring size. Since only THE SGA uses large pages, other database components cannot use them. Should the KVM server and virtual machine use the same memory ratio for large pages? For example, use 50% of the total memory size to avoid imbalance caused by improper proportion. 原始邮件 发件人: Strahil Nikolov 日期: 2021年10月1日周五 22:52 收件人: tommy sway , Strahil Nikolov via Users 主 题: Re: [ovirt-users]Re: 回复: Re: About the vm memory limit Actually yes - should speedup the process. I've edited that draft too many times :) On Fri, Oct 1, 2021 at 7:32, tommy sway wrote: ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/LMPJISCMPKRWJN4KW5DVVWEXLRK45L7L/ ___ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/76DJKE4T7ZVRCOTUS6OEOVNZHJNYCJ3B/