From patchwork Tue Jan 26 08:31:49 2010 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Liran Schour X-Patchwork-Id: 43682 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from lists.gnu.org (lists.gnu.org [199.232.76.165]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTPS id B4860B6F08 for ; Tue, 26 Jan 2010 19:52:36 +1100 (EST) Received: from localhost ([127.0.0.1]:41879 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1NZh6y-0001Rk-Ll for incoming@patchwork.ozlabs.org; Tue, 26 Jan 2010 03:49:24 -0500 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1NZgrM-0003KT-6L for qemu-devel@nongnu.org; Tue, 26 Jan 2010 03:33:17 -0500 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1NZgrE-0003H3-8e for qemu-devel@nongnu.org; Tue, 26 Jan 2010 03:33:13 -0500 Received: from [199.232.76.173] (port=50318 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1NZgrA-0003GT-QK for qemu-devel@nongnu.org; Tue, 26 Jan 2010 03:33:06 -0500 Received: from mx20.gnu.org ([199.232.41.8]:43117) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1NZgrA-0008Jb-EL for qemu-devel@nongnu.org; Tue, 26 Jan 2010 03:33:04 -0500 Received: from mtagate5.uk.ibm.com ([194.196.100.165]) by mx20.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1NZgr8-0003dg-Oi for qemu-devel@nongnu.org; Tue, 26 Jan 2010 03:33:03 -0500 Received: from d06nrmr1806.portsmouth.uk.ibm.com (d06nrmr1806.portsmouth.uk.ibm.com [9.149.39.193]) by mtagate5.uk.ibm.com (8.13.1/8.13.1) with ESMTP id o0Q8X19n019136 for ; Tue, 26 Jan 2010 08:33:01 GMT Received: from d06av01.portsmouth.uk.ibm.com (d06av01.portsmouth.uk.ibm.com [9.149.37.212]) by d06nrmr1806.portsmouth.uk.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id o0Q8X1qt1360082 for ; Tue, 26 Jan 2010 08:33:01 GMT Received: from d06av01.portsmouth.uk.ibm.com (loopback [127.0.0.1]) by d06av01.portsmouth.uk.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id o0Q8X0RT012383 for ; Tue, 26 Jan 2010 08:33:01 GMT Received: from localhost.localdomain (vandev1.haifa.ibm.com [9.148.8.11]) by d06av01.portsmouth.uk.ibm.com (8.12.11.20060308/8.12.11) with ESMTP id o0Q8WvQ8012257; Tue, 26 Jan 2010 08:33:00 GMT From: Liran Schour To: qemu-devel@nongnu.org Date: Tue, 26 Jan 2010 10:31:49 +0200 Message-Id: <1264494709-31848-6-git-send-email-lirans@il.ibm.com> X-Mailer: git-send-email 1.6.0.4 In-Reply-To: <1264494709-31848-5-git-send-email-lirans@il.ibm.com> References: <1264494709-31848-1-git-send-email-lirans@il.ibm.com> <1264494709-31848-2-git-send-email-lirans@il.ibm.com> <1264494709-31848-3-git-send-email-lirans@il.ibm.com> <1264494709-31848-4-git-send-email-lirans@il.ibm.com> <1264494709-31848-5-git-send-email-lirans@il.ibm.com> X-detected-operating-system: by mx20.gnu.org: GNU/Linux 2.6, seldom 2.4 (older, 4) X-detected-operating-system: by monty-python.gnu.org: GNU/Linux 2.6, seldom 2.4 (older, 4) Cc: Liran Schour Subject: [Qemu-devel] [PATCH v3 5/5] Try not to exceed max downtime on stage3 X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Move to stage3 only when remaining work can be done below max downtime. Use qemu_get_clock_ns for measuring read performance. Signed-off-by: Liran Schour --- block-migration.c | 79 +++++++++++++++++++++++++++++++++++++++++++++++------ 1 files changed, 70 insertions(+), 9 deletions(-) diff --git a/block-migration.c b/block-migration.c index d8755d1..0e63596 100644 --- a/block-migration.c +++ b/block-migration.c @@ -15,8 +15,10 @@ #include "block_int.h" #include "hw/hw.h" #include "qemu-queue.h" +#include "qemu-timer.h" #include "monitor.h" #include "block-migration.h" +#include "migration.h" #include #define BLOCK_SIZE (BDRV_SECTORS_PER_DIRTY_CHUNK << BDRV_SECTOR_BITS) @@ -57,6 +59,7 @@ typedef struct BlkMigBlock { QEMUIOVector qiov; BlockDriverAIOCB *aiocb; int ret; + int64_t time; QSIMPLEQ_ENTRY(BlkMigBlock) entry; } BlkMigBlock; @@ -71,7 +74,8 @@ typedef struct BlkMigState { int64_t total_sector_sum; int prev_progress; int bulk_completed; - int dirty_iterations; + long double total_time; + int reads; } BlkMigState; static BlkMigState block_mig_state; @@ -124,12 +128,28 @@ uint64_t blk_mig_bytes_total(void) return sum << BDRV_SECTOR_BITS; } +static inline void add_avg_read_time(int64_t time) +{ + block_mig_state.reads++; + block_mig_state.total_time += time; +} + +static inline long double compute_read_bwidth(void) +{ + assert(block_mig_state.total_time != 0); + return (block_mig_state.reads * BLOCK_SIZE)/ block_mig_state.total_time; +} + static void blk_mig_read_cb(void *opaque, int ret) { BlkMigBlock *blk = opaque; blk->ret = ret; + blk->time = qemu_get_clock_ns(rt_clock) - blk->time; + + add_avg_read_time(blk->time); + QSIMPLEQ_INSERT_TAIL(&block_mig_state.blk_list, blk, entry); block_mig_state.submitted--; @@ -179,6 +199,8 @@ static int mig_save_device_bulk(Monitor *mon, QEMUFile *f, blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE; qemu_iovec_init_external(&blk->qiov, &blk->iov, 1); + blk->time = qemu_get_clock_ns(rt_clock); + blk->aiocb = bdrv_aio_readv(bs, cur_sector, &blk->qiov, nr_sectors, blk_mig_read_cb, blk); if (!blk->aiocb) { @@ -220,6 +242,8 @@ static void init_blk_migration(Monitor *mon, QEMUFile *f) block_mig_state.total_sector_sum = 0; block_mig_state.prev_progress = -1; block_mig_state.bulk_completed = 0; + block_mig_state.total_time = 0; + block_mig_state.reads = 0; for (bs = bdrv_first; bs != NULL; bs = bs->next) { if (bs->type == BDRV_TYPE_HD) { @@ -314,11 +338,13 @@ static int mig_save_device_dirty(Monitor *mon, QEMUFile *f, blk->bmds = bmds; blk->sector = sector; - if(is_async) { + if (is_async) { blk->iov.iov_base = blk->buf; blk->iov.iov_len = nr_sectors * BDRV_SECTOR_SIZE; qemu_iovec_init_external(&blk->qiov, &blk->iov, 1); + blk->time = qemu_get_clock_ns(rt_clock); + blk->aiocb = bdrv_aio_readv(bmds->bs, sector, &blk->qiov, nr_sectors, blk_mig_read_cb, blk); if (!blk->aiocb) { @@ -345,7 +371,7 @@ static int mig_save_device_dirty(Monitor *mon, QEMUFile *f, return (bmds->cur_dirty >= bmds->total_sectors); - error: +error: monitor_printf(mon, "Error reading sector %" PRId64 "\n", sector); qemu_file_set_error(f); qemu_free(blk->buf); @@ -359,7 +385,7 @@ static int blk_mig_save_dirty_block(Monitor *mon, QEMUFile *f, int is_async) int ret = 0; QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) { - if(mig_save_device_dirty(mon, f, bmds, is_async) == 0) { + if (mig_save_device_dirty(mon, f, bmds, is_async) == 0) { ret = 1; break; } @@ -400,9 +426,42 @@ static void flush_blks(QEMUFile* f) block_mig_state.transferred); } +static int64_t get_remaining_dirty(void) +{ + BlkMigDevState *bmds; + int64_t dirty = 0; + + QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) { + dirty += bdrv_get_dirty_count(bmds->bs); + } + + return dirty * BLOCK_SIZE; +} + static int is_stage2_completed(void) { - return (block_mig_state.submitted == 0 && block_mig_state.bulk_completed); + int64_t remaining_dirty; + long double bwidth; + + if (block_mig_state.bulk_completed == 1) { + + remaining_dirty = get_remaining_dirty(); + if (remaining_dirty == 0) { + return 1; + } + + bwidth = compute_read_bwidth(); + + if ((remaining_dirty / bwidth) <= + migrate_max_downtime()) { + /* finish stage2 because we think that we can finish remaing work + below max_downtime */ + + return 1; + } + } + + return 0; } static void blk_mig_cleanup(Monitor *mon) @@ -458,7 +517,7 @@ static int block_save_live(Monitor *mon, QEMUFile *f, int stage, void *opaque) blk_mig_reset_dirty_cursor(); - if(stage == 2) { + if (stage == 2) { /* control the rate of transfer */ while ((block_mig_state.submitted + block_mig_state.read_done) * BLOCK_SIZE < @@ -466,7 +525,7 @@ static int block_save_live(Monitor *mon, QEMUFile *f, int stage, void *opaque) if (block_mig_state.bulk_completed == 0) { /* first finish the bulk phase */ if (blk_mig_save_bulked_block(mon, f) == 0) { - /* finish saving bulk on all devices */ + /* finished saving bulk on all devices */ block_mig_state.bulk_completed = 1; } } else { @@ -486,9 +545,11 @@ static int block_save_live(Monitor *mon, QEMUFile *f, int stage, void *opaque) } if (stage == 3) { - /* we know for sure that save bulk is completed */ + /* we know for sure that save bulk is completed and + all async read completed */ + assert(block_mig_state.submitted == 0); - while(blk_mig_save_dirty_block(mon, f, 0) != 0); + while (blk_mig_save_dirty_block(mon, f, 0) != 0); blk_mig_cleanup(mon); /* report completion */