From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:51264) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Um5hm-00037m-CC for qemu-devel@nongnu.org; Mon, 10 Jun 2013 13:16:39 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Um5hf-0007Cr-UJ for qemu-devel@nongnu.org; Mon, 10 Jun 2013 13:16:30 -0400 Received: from e7.ny.us.ibm.com ([32.97.182.137]:59623) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Um5hf-0007CZ-Nl for qemu-devel@nongnu.org; Mon, 10 Jun 2013 13:16:23 -0400 Received: from /spool/local by e7.ny.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Mon, 10 Jun 2013 13:16:22 -0400 Received: from d01relay03.pok.ibm.com (d01relay03.pok.ibm.com [9.56.227.235]) by d01dlp03.pok.ibm.com (Postfix) with ESMTP id D3418C90058 for ; Mon, 10 Jun 2013 13:16:17 -0400 (EDT) Received: from d01av01.pok.ibm.com (d01av01.pok.ibm.com [9.56.224.215]) by d01relay03.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id r5AHGI9p287308 for ; Mon, 10 Jun 2013 13:16:18 -0400 Received: from d01av01.pok.ibm.com (loopback [127.0.0.1]) by d01av01.pok.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id r5AHGHC5032136 for ; Mon, 10 Jun 2013 13:16:18 -0400 From: mrhines@linux.vnet.ibm.com Date: Mon, 10 Jun 2013 13:16:02 -0400 Message-Id: <1370884574-30057-1-git-send-email-mrhines@linux.vnet.ibm.com> Subject: [Qemu-devel] [PATCH RESEND v7 00/12] rdma: migration support List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: aliguori@us.ibm.com, quintela@redhat.com, knoel@redhat.com, owasserm@redhat.com, abali@us.ibm.com, mrhines@us.ibm.com, gokul@us.ibm.com, pbonzini@redhat.com, chegu_vinod@hp.com From: "Michael R. Hines" Ooops. Forgot the signed-off bys, resending. =) Changes since v6: - added signed-off bys - *Major* bug testing: - In-depth regression testing, several bug fixes. - Over 1000+ back-to-back migrations performed without error. - rdma: introduce qemu_update_position(), since RDMA writes are asynchronous, the accounting needs to be also. - fixed configure when RDMA libraries not available Wiki: http://wiki.qemu.org/Features/RDMALiveMigration Github: git@github.com:hinesmr/qemu.git Here is a brief summary of total migration time and downtime using RDMA: Using a 40gbps infiniband link performing a worst-case stress test, using an 8GB RAM virtual machine: Using the following command: $ apt-get install stress $ stress --vm-bytes 7500M --vm 1 --vm-keep RESULTS: 1. Migration throughput: 26 gigabits/second. 2. Downtime (stop time) varies between 15 and 100 milliseconds. EFFECTS of memory registration on bulk phase round: For example, in the same 8GB RAM example with all 8GB of memory in active use and the VM itself is completely idle using the same 40 gbps infiniband link: 1. x-rdma-pin-all disabled total time: approximately 7.5 seconds @ 9.5 Gbps 2. x-rdma-pin-all enabled total time: approximately 4 seconds @ 26 Gbps These numbers would of course scale up to whatever size virtual machine you have to migrate using RDMA. Enabling this feature does *not* have any measurable affect on migration *downtime*. This is because, without this feature, all of the memory will have already been registered already in advance during the bulk round and does not need to be re-registered during the successive iteration rounds. The following changes since commit 4f293bd6e53739e089f33b458f70a9c4ac136b92: xilinx_axidma: Do not set DMA .notify to NULL after notify (2013-06-10 13:04:40 +0200) are available in the git repository at: git@github.com:hinesmr/qemu.git rdma_patch_v7 for you to fetch changes up to 3f500e47dd4e112ed3e65cb33b6efa5c0f03f0ff: rdma: send pc.ram (2013-06-10 11:52:20 -0400) ---------------------------------------------------------------- Michael R. Hines (12): rdma: add documentation rdma: introduce qemu_update_position() rdma: export yield_until_fd_readable() rdma: export throughput w/ MigrationStats QMP rdma: introduce qemu_file_mode_is_not_valid() rdma: export qemu_fflush() rdma: introduce ram_handle_compressed() rdma: introduce qemu_ram_foreach_block() rdma: new QEMUFileOps hooks rdma: introduce capability x-rdma-pin-all rdma: core logic rdma: send pc.ram Makefile.objs | 1 + arch_init.c | 74 +- configure | 29 + docs/rdma.txt | 404 ++++++ exec.c | 9 + hmp.c | 2 + include/block/coroutine.h | 6 + include/exec/cpu-common.h | 5 + include/migration/migration.h | 31 + include/migration/qemu-file.h | 32 + migration-rdma.c | 2809 +++++++++++++++++++++++++++++++++++++++++ migration.c | 23 + qapi-schema.json | 12 +- qemu-coroutine-io.c | 23 + savevm.c | 114 +- 15 files changed, 3527 insertions(+), 47 deletions(-) create mode 100644 docs/rdma.txt create mode 100644 migration-rdma.c -- 1.7.10.4