From patchwork Mon Feb 24 02:35:40 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Keqian Zhu X-Patchwork-Id: 1242790 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=) Authentication-Results: ozlabs.org; dmarc=none (p=none dis=none) header.from=huawei.com Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 48QmTB0d5Xz9sNg for ; Mon, 24 Feb 2020 13:38:22 +1100 (AEDT) Received: from localhost ([::1]:58922 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1j63dU-0005BO-2G for incoming@patchwork.ozlabs.org; Sun, 23 Feb 2020 21:38:20 -0500 Received: from eggs.gnu.org ([2001:470:142:3::10]:53316) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1j63d0-00055I-9L for qemu-devel@nongnu.org; Sun, 23 Feb 2020 21:37:51 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1j63cy-0007Z7-CQ for qemu-devel@nongnu.org; Sun, 23 Feb 2020 21:37:50 -0500 Received: from szxga04-in.huawei.com ([45.249.212.190]:2722 helo=huawei.com) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1j63cu-0007YV-5d; Sun, 23 Feb 2020 21:37:44 -0500 Received: from DGGEMS402-HUB.china.huawei.com (unknown [172.30.72.60]) by Forcepoint Email with ESMTP id 8AD8DA45FC4AB7FA7B02; Mon, 24 Feb 2020 10:37:41 +0800 (CST) Received: from linux-TFkxOR.huawei.com (10.175.104.212) by DGGEMS402-HUB.china.huawei.com (10.3.19.202) with Microsoft SMTP Server id 14.3.439.0; Mon, 24 Feb 2020 10:37:31 +0800 From: Keqian Zhu To: Subject: [RESEND PATCH v2] migration/throttle: Add throttle-trig-thres migration parameter Date: Mon, 24 Feb 2020 10:35:40 +0800 Message-ID: <20200224023540.15147-1-zhukeqian1@huawei.com> X-Mailer: git-send-email 2.19.1 MIME-Version: 1.0 X-Originating-IP: [10.175.104.212] X-CFilter-Loop: Reflected X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] [fuzzy] X-Received-From: 45.249.212.190 X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Juan Quintela , "Dr. David Alan Gilbert" , Markus Armbruster , qemu-arm@nongnu.org, wanghaibin.wang@huawei.com, Keqian Zhu Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: "Qemu-devel" Currently, if the bytes_dirty_period is more than the 50% of bytes_xfer_period, we start or increase throttling. If we make this percentage higher, then we can tolerate higher dirty rate during migration, which means less impact on guest. The side effect of higher percentage is longer migration time. We can make this parameter configurable to switch between mig- ration time first or guest performance first. The default value is 50 and valid range is 1 to 100. Signed-off-by: Keqian Zhu --- Changelog: v1->v2 -Use full name for parameter. Suggested by Eric Blake. -Change the upper bound of threshold to 100. -Extract the throttle strategy as function. --- Cc: Juan Quintela Cc: "Dr. David Alan Gilbert" Cc: Eric Blake Cc: Markus Armbruster --- migration/migration.c | 24 ++++++++++++++++++++ migration/ram.c | 52 +++++++++++++++++++++++++------------------ monitor/hmp-cmds.c | 7 ++++++ qapi/migration.json | 16 ++++++++++++- 4 files changed, 76 insertions(+), 23 deletions(-) diff --git a/migration/migration.c b/migration/migration.c index 8fb68795dc..42d2d556e3 100644 --- a/migration/migration.c +++ b/migration/migration.c @@ -78,6 +78,7 @@ /*0: means nocompress, 1: best speed, ... 9: best compress ratio */ #define DEFAULT_MIGRATE_COMPRESS_LEVEL 1 /* Define default autoconverge cpu throttle migration parameters */ +#define DEFAULT_MIGRATE_THROTTLE_TRIGGER_THRESHOLD 50 #define DEFAULT_MIGRATE_CPU_THROTTLE_INITIAL 20 #define DEFAULT_MIGRATE_CPU_THROTTLE_INCREMENT 10 #define DEFAULT_MIGRATE_MAX_CPU_THROTTLE 99 @@ -778,6 +779,8 @@ MigrationParameters *qmp_query_migrate_parameters(Error **errp) params->compress_wait_thread = s->parameters.compress_wait_thread; params->has_decompress_threads = true; params->decompress_threads = s->parameters.decompress_threads; + params->has_throttle_trigger_threshold = true; + params->throttle_trigger_threshold = s->parameters.throttle_trigger_threshold; params->has_cpu_throttle_initial = true; params->cpu_throttle_initial = s->parameters.cpu_throttle_initial; params->has_cpu_throttle_increment = true; @@ -1164,6 +1167,15 @@ static bool migrate_params_check(MigrationParameters *params, Error **errp) return false; } + if (params->has_throttle_trigger_threshold && + (params->throttle_trigger_threshold < 1 || + params->throttle_trigger_threshold > 100)) { + error_setg(errp, QERR_INVALID_PARAMETER_VALUE, + "throttle_trigger_threshold", + "an integer in the range of 1 to 100"); + return false; + } + if (params->has_cpu_throttle_initial && (params->cpu_throttle_initial < 1 || params->cpu_throttle_initial > 99)) { @@ -1279,6 +1291,10 @@ static void migrate_params_test_apply(MigrateSetParameters *params, dest->decompress_threads = params->decompress_threads; } + if (params->has_throttle_trigger_threshold) { + dest->throttle_trigger_threshold = params->throttle_trigger_threshold; + } + if (params->has_cpu_throttle_initial) { dest->cpu_throttle_initial = params->cpu_throttle_initial; } @@ -1360,6 +1376,10 @@ static void migrate_params_apply(MigrateSetParameters *params, Error **errp) s->parameters.decompress_threads = params->decompress_threads; } + if (params->has_throttle_trigger_threshold) { + s->parameters.throttle_trigger_threshold = params->throttle_trigger_threshold; + } + if (params->has_cpu_throttle_initial) { s->parameters.cpu_throttle_initial = params->cpu_throttle_initial; } @@ -3506,6 +3526,9 @@ static Property migration_properties[] = { DEFINE_PROP_UINT8("x-decompress-threads", MigrationState, parameters.decompress_threads, DEFAULT_MIGRATE_DECOMPRESS_THREAD_COUNT), + DEFINE_PROP_UINT8("x-throttle-trigger-threshold", MigrationState, + parameters.throttle_trigger_threshold, + DEFAULT_MIGRATE_THROTTLE_TRIGGER_THRESHOLD), DEFINE_PROP_UINT8("x-cpu-throttle-initial", MigrationState, parameters.cpu_throttle_initial, DEFAULT_MIGRATE_CPU_THROTTLE_INITIAL), @@ -3606,6 +3629,7 @@ static void migration_instance_init(Object *obj) params->has_compress_level = true; params->has_compress_threads = true; params->has_decompress_threads = true; + params->has_throttle_trigger_threshold = true; params->has_cpu_throttle_initial = true; params->has_cpu_throttle_increment = true; params->has_max_bandwidth = true; diff --git a/migration/ram.c b/migration/ram.c index ed23ed1c7c..3a38253903 100644 --- a/migration/ram.c +++ b/migration/ram.c @@ -896,11 +896,38 @@ static void migration_update_rates(RAMState *rs, int64_t end_time) } } +static void migration_trigger_throttle(RAMState *rs) +{ + MigrationState *s = migrate_get_current(); + uint64_t threshold = s->parameters.throttle_trigger_threshold; + + uint64_t bytes_xfer_period = ram_counters.transferred - rs->bytes_xfer_prev; + uint64_t bytes_dirty_period = rs->num_dirty_pages_period * TARGET_PAGE_SIZE; + uint64_t bytes_dirty_threshold = bytes_xfer_period * threshold / 100; + + /* During block migration the auto-converge logic incorrectly detects + * that ram migration makes no progress. Avoid this by disabling the + * throttling logic during the bulk phase of block migration. */ + if (migrate_auto_converge() && !blk_mig_bulk_active()) { + /* The following detection logic can be refined later. For now: + Check to see if the ratio between dirtied bytes and the approx. + amount of bytes that just got transferred since the last time + we were in this routine reaches the threshold. If that happens + twice, start or increase throttling. */ + + if ((bytes_dirty_period > bytes_dirty_threshold) && + (++rs->dirty_rate_high_cnt >= 2)) { + trace_migration_throttle(); + rs->dirty_rate_high_cnt = 0; + mig_throttle_guest_down(); + } + } +} + static void migration_bitmap_sync(RAMState *rs) { RAMBlock *block; int64_t end_time; - uint64_t bytes_xfer_now; ram_counters.dirty_sync_count++; @@ -927,26 +954,7 @@ static void migration_bitmap_sync(RAMState *rs) /* more than 1 second = 1000 millisecons */ if (end_time > rs->time_last_bitmap_sync + 1000) { - bytes_xfer_now = ram_counters.transferred; - - /* During block migration the auto-converge logic incorrectly detects - * that ram migration makes no progress. Avoid this by disabling the - * throttling logic during the bulk phase of block migration. */ - if (migrate_auto_converge() && !blk_mig_bulk_active()) { - /* The following detection logic can be refined later. For now: - Check to see if the dirtied bytes is 50% more than the approx. - amount of bytes that just got transferred since the last time we - were in this routine. If that happens twice, start or increase - throttling */ - - if ((rs->num_dirty_pages_period * TARGET_PAGE_SIZE > - (bytes_xfer_now - rs->bytes_xfer_prev) / 2) && - (++rs->dirty_rate_high_cnt >= 2)) { - trace_migration_throttle(); - rs->dirty_rate_high_cnt = 0; - mig_throttle_guest_down(); - } - } + migration_trigger_throttle(rs); migration_update_rates(rs, end_time); @@ -955,7 +963,7 @@ static void migration_bitmap_sync(RAMState *rs) /* reset period counters */ rs->time_last_bitmap_sync = end_time; rs->num_dirty_pages_period = 0; - rs->bytes_xfer_prev = bytes_xfer_now; + rs->bytes_xfer_prev = ram_counters.transferred; } if (migrate_use_events()) { qapi_event_send_migration_pass(ram_counters.dirty_sync_count); diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c index 53bc3f76c4..de67d0bd53 100644 --- a/monitor/hmp-cmds.c +++ b/monitor/hmp-cmds.c @@ -409,6 +409,10 @@ void hmp_info_migrate_parameters(Monitor *mon, const QDict *qdict) monitor_printf(mon, "%s: %u\n", MigrationParameter_str(MIGRATION_PARAMETER_DECOMPRESS_THREADS), params->decompress_threads); + assert(params->has_throttle_trigger_threshold); + monitor_printf(mon, "%s: %u\n", + MigrationParameter_str(MIGRATION_PARAMETER_THROTTLE_TRIGGER_THRESHOLD), + params->throttle_trigger_threshold); assert(params->has_cpu_throttle_initial); monitor_printf(mon, "%s: %u\n", MigrationParameter_str(MIGRATION_PARAMETER_CPU_THROTTLE_INITIAL), @@ -1764,6 +1768,9 @@ void hmp_migrate_set_parameter(Monitor *mon, const QDict *qdict) p->has_decompress_threads = true; visit_type_int(v, param, &p->decompress_threads, &err); break; + case MIGRATION_PARAMETER_THROTTLE_TRIGGER_THRESHOLD: + p->has_throttle_trigger_threshold = true; + visit_type_int(v, param, &p->throttle_trigger_threshold, &err); case MIGRATION_PARAMETER_CPU_THROTTLE_INITIAL: p->has_cpu_throttle_initial = true; visit_type_int(v, param, &p->cpu_throttle_initial, &err); diff --git a/qapi/migration.json b/qapi/migration.json index 52f3429969..0e7ac64c98 100644 --- a/qapi/migration.json +++ b/qapi/migration.json @@ -524,6 +524,10 @@ # compression, so set the decompress-threads to the number about 1/4 # of compress-threads is adequate. # +# @throttle-trigger-threshold: The ratio of bytes_dirty_period and bytes_xfer_period +# to trigger throttling. It is expressed as percentage. +# The default value is 50. (Since 5.0) +# # @cpu-throttle-initial: Initial percentage of time guest cpus are throttled # when migration auto-converge is activated. The # default value is 20. (Since 2.7) @@ -592,7 +596,7 @@ 'data': ['announce-initial', 'announce-max', 'announce-rounds', 'announce-step', 'compress-level', 'compress-threads', 'decompress-threads', - 'compress-wait-thread', + 'compress-wait-thread', 'throttle-trigger-threshold', 'cpu-throttle-initial', 'cpu-throttle-increment', 'tls-creds', 'tls-hostname', 'tls-authz', 'max-bandwidth', 'downtime-limit', 'x-checkpoint-delay', 'block-incremental', @@ -626,6 +630,10 @@ # # @decompress-threads: decompression thread count # +# @throttle-trigger-threshold: The ratio of bytes_dirty_period and bytes_xfer_period +# to trigger throttling. It is expressed as percentage. +# The default value is 50. (Since 5.0) +# # @cpu-throttle-initial: Initial percentage of time guest cpus are # throttled when migration auto-converge is activated. # The default value is 20. (Since 2.7) @@ -701,6 +709,7 @@ '*compress-threads': 'int', '*compress-wait-thread': 'bool', '*decompress-threads': 'int', + '*throttle-trigger-threshold': 'int', '*cpu-throttle-initial': 'int', '*cpu-throttle-increment': 'int', '*tls-creds': 'StrOrNull', @@ -759,6 +768,10 @@ # # @decompress-threads: decompression thread count # +# @throttle-trigger-threshold: The ratio of bytes_dirty_period and bytes_xfer_period +# to trigger throttling. It is expressed as percentage. +# The default value is 50. (Since 5.0) +# # @cpu-throttle-initial: Initial percentage of time guest cpus are # throttled when migration auto-converge is activated. # (Since 2.7) @@ -834,6 +847,7 @@ '*compress-threads': 'uint8', '*compress-wait-thread': 'bool', '*decompress-threads': 'uint8', + '*throttle-trigger-threshold': 'uint8', '*cpu-throttle-initial': 'uint8', '*cpu-throttle-increment': 'uint8', '*tls-creds': 'str',