From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Jan Beulich" Subject: Re: Xen Platform QoS design discussion Date: Thu, 29 May 2014 08:01:40 +0100 Message-ID: <5386E96402000078000B525D@mail.emea.novell.com> References: <40776A41FC278F40B59438AD47D147A9119F3FEA@SHSMSX104.ccr.corp.intel.com> <5363804B020000780000E604@mail.emea.novell.com> <40776A41FC278F40B59438AD47D147A9119F4EF4@SHSMSX104.ccr.corp.intel.com> <5363AE54020000780000E7A2@mail.emea.novell.com> <40776A41FC278F40B59438AD47D147A9119FE6BB@SHSMSX104.ccr.corp.intel.com> <5368B418.9000307@citrix.com> <536AA342.8030003@citrix.com> <40776A41FC278F40B59438AD47D147A911A00A4C@SHSMSX104.ccr.corp.intel.com> <536B69AB.7010005@citrix.com> <40776A41FC278F40B59438AD47D147A911A150FC@SHSMSX104.ccr.corp.intel.com> <537A0B17020000780001390F@mail.emea.novell.com> <5379F576.4050108@eu.citrix.com> <537A18260200007800013A06@mail.emea.novell.com> <40776A41FC278F40B59438AD47D147A911A1AAEC@SHSMSX104.ccr.corp.intel.com> <537DD3E60200007800014CFD@mail.emea.novell.com> <537DC2F2.30702@eu.citrix.com> <40776A41FC278F40B59438AD47D147A911A206A4@SHSMSX104.ccr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <40776A41FC278F40B59438AD47D147A911A206A4@SHSMSX104.ccr.corp.intel.com> Content-Disposition: inline List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: george.dunlap@eu.citrix.com, dongxiao.xu@intel.com Cc: andrew.cooper3@citrix.com, Ian.Campbell@citrix.com, xen-devel@lists.xen.org List-Id: xen-devel@lists.xenproject.org >>> "Xu, Dongxiao" 05/29/14 2:46 AM >>> >I think Jan's opinion here is similar to what I proposed in the beginning of this thread. >The only difference is that, Jan prefers to get the CQM data per-socket and per-domain >with data copying, while I proposed to get the CQM data per-domain for all sockets >that can reduce the amount of hypercalls. I don't think I ever voiced any preference between these two. All I said it depends on prevalent usage models, and to date I don't think I've seen a proper analysis of what the main usage model would be - it all seems guesswork and/or taking random examples. What I did say I'd prefer is to have all this done outside the hypervisor, with the hypervisor just providing fundamental infrastructure (MSR accesses). Jan