|
From: | zhanghailiang |
Subject: | Re: [Qemu-devel] [PATCH COLO-Frame v9 00/32] COarse-grain LOck-stepping(COLO) Virtual Machines for Non-stop Service (FT) |
Date: | Thu, 22 Oct 2015 17:01:23 +0800 |
User-agent: | Mozilla/5.0 (Windows NT 6.1; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.1.0 |
Hi Dave, Thanks for feedback, i have recorded this problem, maybe we can optimize this calling in the future. Thanks, zhanghailiang On 2015/10/21 22:10, Dr. David Alan Gilbert wrote:
Hi, Not something I found on this set; but on your older set, where you have the experimental 'live' ram transfer; we found a test that was slowed down (5x) by the calls to 'colo_need_live_migrate_ram'; even if it decides not to send any RAM in the live phase, the call to 'qemu_savevm_state_pending' can be really expensive. Just thought it worth mentioning. Dave -- Dr. David Alan Gilbert / address@hidden / Manchester, UK .
[Prev in Thread] | Current Thread | [Next in Thread] |