From patchwork Mon Mar 1 13:04:26 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Anton Ivanov X-Patchwork-Id: 1445563 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=openvswitch.org (client-ip=140.211.166.133; helo=smtp2.osuosl.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Received: from smtp2.osuosl.org (smtp2.osuosl.org [140.211.166.133]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 4Dq0pm20c1z9sRR for ; Tue, 2 Mar 2021 00:04:48 +1100 (AEDT) Received: from localhost (localhost [127.0.0.1]) by smtp2.osuosl.org (Postfix) with ESMTP id B75084306E; Mon, 1 Mar 2021 13:04:45 +0000 (UTC) X-Virus-Scanned: amavisd-new at osuosl.org Received: from smtp2.osuosl.org ([127.0.0.1]) by localhost (smtp2.osuosl.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id mQUloXiYQp3Q; Mon, 1 Mar 2021 13:04:44 +0000 (UTC) Received: from lists.linuxfoundation.org (lf-lists.osuosl.org [140.211.9.56]) by smtp2.osuosl.org (Postfix) with ESMTP id 6CF1643063; Mon, 1 Mar 2021 13:04:43 +0000 (UTC) Received: from lf-lists.osuosl.org (localhost [127.0.0.1]) by lists.linuxfoundation.org (Postfix) with ESMTP id B066CC0011; Mon, 1 Mar 2021 13:04:41 +0000 (UTC) X-Original-To: ovs-dev@openvswitch.org Delivered-To: ovs-dev@lists.linuxfoundation.org Received: from smtp2.osuosl.org (smtp2.osuosl.org [140.211.166.133]) by lists.linuxfoundation.org (Postfix) with ESMTP id 4016CC0001 for ; Mon, 1 Mar 2021 13:04:40 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by smtp2.osuosl.org (Postfix) with ESMTP id D304843062 for ; Mon, 1 Mar 2021 13:04:39 +0000 (UTC) X-Virus-Scanned: amavisd-new at osuosl.org Received: from smtp2.osuosl.org ([127.0.0.1]) by localhost (smtp2.osuosl.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id AduCSHJCRm_s for ; Mon, 1 Mar 2021 13:04:38 +0000 (UTC) X-Greylist: from auto-whitelisted by SQLgrey-1.8.0 Received: from www.kot-begemot.co.uk (ivanoab7.miniserver.com [37.128.132.42]) by smtp2.osuosl.org (Postfix) with ESMTPS id 1818E4306B for ; Mon, 1 Mar 2021 13:04:37 +0000 (UTC) Received: from tun252.jain.kot-begemot.co.uk ([192.168.18.6] helo=jain.kot-begemot.co.uk) by www.kot-begemot.co.uk with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1lGiDz-00072L-I7; Mon, 01 Mar 2021 13:04:35 +0000 Received: from jain.kot-begemot.co.uk ([192.168.3.3]) by jain.kot-begemot.co.uk with esmtp (Exim 4.92) (envelope-from ) id 1lGiDw-0005VP-Bm; Mon, 01 Mar 2021 13:04:34 +0000 From: anton.ivanov@cambridgegreys.com To: ovs-dev@openvswitch.org Date: Mon, 1 Mar 2021 13:04:26 +0000 Message-Id: <20210301130427.21069-2-anton.ivanov@cambridgegreys.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20210301130427.21069-1-anton.ivanov@cambridgegreys.com> References: <20210301130427.21069-1-anton.ivanov@cambridgegreys.com> MIME-Version: 1.0 X-Clacks-Overhead: GNU Terry Pratchett Cc: i.maximets@ovn.org, Anton Ivanov Subject: [ovs-dev] [OVN Patch v15 2/3] ovn-northd: Introduce parallel lflow build X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: ovs-dev-bounces@openvswitch.org Sender: "dev" From: Anton Ivanov Datapaths, ports, igmp groups and load balancers can now be iterated over in parallel in order to speed up the lflow generation. This decreases the time needed to generate the logical flows by a factor of 4+ on a 6 core/12 thread CPU without datapath groups - from 0.8-1 microseconds per flow down to 0.2-0.3 microseconds per flow on average. The decrease in time to compute lflows with datapath groups enabled is ~2 times for the same hardware - from an average of 2.4 microseconds per flow to 1.2 microseconds per flow. Tested for on an 8 node, 400 pod K8 simulation resulting in > 6K flows. Signed-off-by: Anton Ivanov --- northd/ovn-northd.c | 325 ++++++++++++++++++++++++++++++++++++-------- 1 file changed, 272 insertions(+), 53 deletions(-) diff --git a/northd/ovn-northd.c b/northd/ovn-northd.c index ac872aade..4b2591385 100644 --- a/northd/ovn-northd.c +++ b/northd/ovn-northd.c @@ -39,6 +39,7 @@ #include "lib/ovn-util.h" #include "lib/lb.h" #include "memory.h" +#include "lib/ovn-parallel-hmap.h" #include "ovn/actions.h" #include "ovn/features.h" #include "ovn/logical-fields.h" @@ -4067,6 +4068,34 @@ ovn_lflow_init(struct ovn_lflow *lflow, struct ovn_datapath *od, /* If this option is 'true' northd will combine logical flows that differ by * logical datapath only by creating a datapath group. */ static bool use_logical_dp_groups = false; +static bool use_parallel_build = true; + +static struct hashrow_locks lflow_locks; + +/* Adds a row with the specified contents to the Logical_Flow table. + * Version to use when locking is required. + */ +static void +do_ovn_lflow_add(struct hmap *lflow_map, bool shared, + struct ovn_datapath *od, + uint32_t hash, struct ovn_lflow *lflow) +{ + + struct ovn_lflow *old_lflow; + + if (shared && use_logical_dp_groups) { + old_lflow = ovn_lflow_find_by_lflow(lflow_map, lflow, hash); + if (old_lflow) { + ovn_lflow_destroy(NULL, lflow); + hmapx_add(&old_lflow->od_group, od); + return; + } + } + + hmapx_add(&lflow->od_group, od); + hmap_insert_fast(lflow_map, &lflow->hmap_node, hash); +} + /* Adds a row with the specified contents to the Logical_Flow table. */ static void @@ -4077,7 +4106,7 @@ ovn_lflow_add_at(struct hmap *lflow_map, struct ovn_datapath *od, { ovs_assert(ovn_stage_to_datapath_type(stage) == ovn_datapath_get_type(od)); - struct ovn_lflow *old_lflow, *lflow; + struct ovn_lflow *lflow; uint32_t hash; lflow = xmalloc(sizeof *lflow); @@ -4089,17 +4118,14 @@ ovn_lflow_add_at(struct hmap *lflow_map, struct ovn_datapath *od, ovn_lflow_hint(stage_hint), where); hash = ovn_lflow_hash(lflow); - if (shared && use_logical_dp_groups) { - old_lflow = ovn_lflow_find_by_lflow(lflow_map, lflow, hash); - if (old_lflow) { - ovn_lflow_destroy(NULL, lflow); - hmapx_add(&old_lflow->od_group, od); - return; - } - } - hmapx_add(&lflow->od_group, od); - hmap_insert(lflow_map, &lflow->hmap_node, hash); + if (use_logical_dp_groups && use_parallel_build) { + lock_hash_row(&lflow_locks, hash); + do_ovn_lflow_add(lflow_map, shared, od, hash, lflow); + unlock_hash_row(&lflow_locks, hash); + } else { + do_ovn_lflow_add(lflow_map, shared, od, hash, lflow); + } } /* Adds a row with the specified contents to the Logical_Flow table. */ @@ -7347,6 +7373,8 @@ build_lswitch_ip_mcast_igmp_mld(struct ovn_igmp_group *igmp_group, } } +static struct ovs_mutex mcgroup_mutex = OVS_MUTEX_INITIALIZER; + /* Ingress table 19: Destination lookup, unicast handling (priority 50), */ static void build_lswitch_ip_unicast_lookup(struct ovn_port *op, @@ -7385,7 +7413,9 @@ build_lswitch_ip_unicast_lookup(struct ovn_port *op, &op->nbsp->header_); } else if (!strcmp(op->nbsp->addresses[i], "unknown")) { if (lsp_is_enabled(op->nbsp)) { + ovs_mutex_lock(&mcgroup_mutex); ovn_multicast_add(mcgroups, &mc_unknown, op); + ovs_mutex_unlock(&mcgroup_mutex); op->od->has_unknown = true; } } else if (is_dynamic_lsp_address(op->nbsp->addresses[i])) { @@ -11757,6 +11787,120 @@ build_lswitch_and_lrouter_iterate_by_op(struct ovn_port *op, &lsi->actions); } +struct lflows_thread_pool { + struct worker_pool *pool; +}; + +static void *build_lflows_thread(void *arg) +{ + struct worker_control *control = (struct worker_control *) arg; + struct lflows_thread_pool *workload; + struct lswitch_flow_build_info *lsi; + + struct ovn_datapath *od; + struct ovn_port *op; + struct ovn_northd_lb *lb; + struct ovn_igmp_group *igmp_group; + int bnum; + + while (!stop_parallel_processing()) { + wait_for_work(control); + workload = (struct lflows_thread_pool *) control->workload; + lsi = (struct lswitch_flow_build_info *) control->data; + if (lsi && workload) { + /* Iterate over bucket ThreadID, ThreadID+size, ... */ + for (bnum = control->id; + bnum <= lsi->datapaths->mask; + bnum += workload->pool->size) + { + HMAP_FOR_EACH_IN_PARALLEL (od, key_node, bnum, lsi->datapaths) { + if (stop_parallel_processing()) { + return NULL; + } + build_lswitch_and_lrouter_iterate_by_od(od, lsi); + } + } + for (bnum = control->id; + bnum <= lsi->ports->mask; + bnum += workload->pool->size) + { + HMAP_FOR_EACH_IN_PARALLEL (op, key_node, bnum, lsi->ports) { + if (stop_parallel_processing()) { + return NULL; + } + build_lswitch_and_lrouter_iterate_by_op(op, lsi); + } + } + for (bnum = control->id; + bnum <= lsi->lbs->mask; + bnum += workload->pool->size) + { + HMAP_FOR_EACH_IN_PARALLEL (lb, hmap_node, bnum, lsi->lbs) { + if (stop_parallel_processing()) { + return NULL; + } + build_lswitch_arp_nd_service_monitor(lb, lsi->lflows, + &lsi->match, + &lsi->actions); + } + } + for (bnum = control->id; + bnum <= lsi->igmp_groups->mask; + bnum += workload->pool->size) + { + HMAP_FOR_EACH_IN_PARALLEL ( + igmp_group, hmap_node, bnum, lsi->igmp_groups) { + if (stop_parallel_processing()) { + return NULL; + } + build_lswitch_ip_mcast_igmp_mld(igmp_group, lsi->lflows, + &lsi->match, + &lsi->actions); + } + } + } + post_completed_work(control); + } + return NULL; +} + +static bool pool_init_done = false; +static struct lflows_thread_pool *build_lflows_pool = NULL; + +static void init_lflows_thread_pool(void) +{ + int index; + + if (!pool_init_done) { + struct worker_pool *pool = add_worker_pool(build_lflows_thread); + pool_init_done = true; + if (pool) { + build_lflows_pool = xmalloc(sizeof(*build_lflows_pool)); + build_lflows_pool->pool = pool; + for (index = 0; index < build_lflows_pool->pool->size; index++) { + build_lflows_pool->pool->controls[index].workload = + build_lflows_pool; + } + } + } +} + +/* TODO: replace hard cutoffs by configurable via commands. These are + * temporary defines to determine single-thread to multi-thread processing + * cutoff. + * Setting to 1 forces "all parallel" lflow build. + */ + +static void +noop_callback(struct worker_pool *pool OVS_UNUSED, + void *fin_result OVS_UNUSED, + void *result_frags OVS_UNUSED, + int index OVS_UNUSED) +{ + /* Do nothing */ +} + + static void build_lswitch_and_lrouter_flows(struct hmap *datapaths, struct hmap *ports, struct hmap *port_groups, struct hmap *lflows, @@ -11765,53 +11909,114 @@ build_lswitch_and_lrouter_flows(struct hmap *datapaths, struct hmap *ports, struct shash *meter_groups, struct hmap *lbs, struct hmap *bfd_connections) { - struct ovn_datapath *od; - struct ovn_port *op; - struct ovn_northd_lb *lb; - struct ovn_igmp_group *igmp_group; char *svc_check_match = xasprintf("eth.dst == %s", svc_monitor_mac); - struct lswitch_flow_build_info lsi = { - .datapaths = datapaths, - .ports = ports, - .port_groups = port_groups, - .lflows = lflows, - .mcgroups = mcgroups, - .igmp_groups = igmp_groups, - .meter_groups = meter_groups, - .lbs = lbs, - .bfd_connections = bfd_connections, - .svc_check_match = svc_check_match, - .match = DS_EMPTY_INITIALIZER, - .actions = DS_EMPTY_INITIALIZER, - }; - - /* Combined build - all lflow generation from lswitch and lrouter - * will move here and will be reogranized by iterator type. - */ - HMAP_FOR_EACH (od, key_node, datapaths) { - build_lswitch_and_lrouter_iterate_by_od(od, &lsi); - } - HMAP_FOR_EACH (op, key_node, ports) { - build_lswitch_and_lrouter_iterate_by_op(op, &lsi); - } - HMAP_FOR_EACH (lb, hmap_node, lbs) { - build_lswitch_arp_nd_service_monitor(lb, lsi.lflows, - &lsi.actions, - &lsi.match); - } - HMAP_FOR_EACH (igmp_group, hmap_node, igmp_groups) { - build_lswitch_ip_mcast_igmp_mld(igmp_group, - lsi.lflows, - &lsi.actions, - &lsi.match); + if (use_parallel_build) { + init_lflows_thread_pool(); + if (!can_parallelize_hashes(false)) { + use_parallel_build = false; + } } - free(svc_check_match); - ds_destroy(&lsi.match); - ds_destroy(&lsi.actions); + if (use_parallel_build) { + struct hmap *lflow_segs; + struct lswitch_flow_build_info *lsiv; + int index; + lsiv = xcalloc(sizeof(*lsiv), build_lflows_pool->pool->size); + if (use_logical_dp_groups) { + lflow_segs = NULL; + } else { + lflow_segs = xcalloc(sizeof(*lflow_segs), build_lflows_pool->pool->size); + } + + /* Set up "work chunks" for each thread to work on. */ + + for (index = 0; index < build_lflows_pool->pool->size; index++) { + if (use_logical_dp_groups) { + /* if dp_groups are in use we lock a shared lflows hash + * on a per-bucket level instead of merging hash frags */ + lsiv[index].lflows = lflows; + } else { + fast_hmap_init(&lflow_segs[index], lflows->mask); + lsiv[index].lflows = &lflow_segs[index]; + } + + lsiv[index].datapaths = datapaths; + lsiv[index].ports = ports; + lsiv[index].port_groups = port_groups; + lsiv[index].mcgroups = mcgroups; + lsiv[index].igmp_groups = igmp_groups; + lsiv[index].meter_groups = meter_groups; + lsiv[index].lbs = lbs; + lsiv[index].bfd_connections = bfd_connections; + lsiv[index].svc_check_match = svc_check_match; + ds_init(&lsiv[index].match); + ds_init(&lsiv[index].actions); + + build_lflows_pool->pool->controls[index].data = &lsiv[index]; + } + + /* Run thread pool. */ + if (use_logical_dp_groups) { + run_pool_callback(build_lflows_pool->pool, NULL, NULL, noop_callback); + } else { + run_pool_hash(build_lflows_pool->pool, lflows, lflow_segs); + } + + for (index = 0; index < build_lflows_pool->pool->size; index++) { + ds_destroy(&lsiv[index].match); + ds_destroy(&lsiv[index].actions); + } + free(lflow_segs); + free(lsiv); + } else { + struct ovn_datapath *od; + struct ovn_port *op; + struct ovn_northd_lb *lb; + struct ovn_igmp_group *igmp_group; + struct lswitch_flow_build_info lsi = { + .datapaths = datapaths, + .ports = ports, + .port_groups = port_groups, + .lflows = lflows, + .mcgroups = mcgroups, + .igmp_groups = igmp_groups, + .meter_groups = meter_groups, + .lbs = lbs, + .bfd_connections = bfd_connections, + .svc_check_match = svc_check_match, + .match = DS_EMPTY_INITIALIZER, + .actions = DS_EMPTY_INITIALIZER, + }; + + /* Combined build - all lflow generation from lswitch and lrouter + * will move here and will be reogranized by iterator type. + */ + HMAP_FOR_EACH (od, key_node, datapaths) { + build_lswitch_and_lrouter_iterate_by_od(od, &lsi); + } + HMAP_FOR_EACH (op, key_node, ports) { + build_lswitch_and_lrouter_iterate_by_op(op, &lsi); + } + HMAP_FOR_EACH (lb, hmap_node, lbs) { + build_lswitch_arp_nd_service_monitor(lb, lsi.lflows, + &lsi.actions, + &lsi.match); + } + HMAP_FOR_EACH (igmp_group, hmap_node, igmp_groups) { + build_lswitch_ip_mcast_igmp_mld(igmp_group, + lsi.lflows, + &lsi.actions, + &lsi.match); + } + + ds_destroy(&lsi.match); + ds_destroy(&lsi.actions); + } + + free(svc_check_match); build_lswitch_flows(datapaths, lflows); } @@ -11882,6 +12087,8 @@ ovn_sb_set_lflow_logical_dp_group( sbrec_logical_flow_set_logical_dp_group(sbflow, dpg->dp_group); } +static ssize_t max_seen_lflow_size = 128; + /* Updates the Logical_Flow and Multicast_Group tables in the OVN_SB database, * constructing their contents based on the OVN_NB database. */ static void @@ -11891,13 +12098,21 @@ build_lflows(struct northd_context *ctx, struct hmap *datapaths, struct shash *meter_groups, struct hmap *lbs, struct hmap *bfd_connections) { - struct hmap lflows = HMAP_INITIALIZER(&lflows); + struct hmap lflows; + fast_hmap_size_for(&lflows, max_seen_lflow_size); + if (use_parallel_build) { + update_hashrow_locks(&lflows, &lflow_locks); + } build_lswitch_and_lrouter_flows(datapaths, ports, port_groups, &lflows, mcgroups, igmp_groups, meter_groups, lbs, bfd_connections); + if (hmap_count(&lflows) > max_seen_lflow_size) { + max_seen_lflow_size = hmap_count(&lflows); + } + /* Collecting all unique datapath groups. */ struct hmap dp_groups = HMAP_INITIALIZER(&dp_groups); struct hmapx single_dp_lflows = HMAPX_INITIALIZER(&single_dp_lflows); @@ -13672,6 +13887,9 @@ main(int argc, char *argv[]) daemonize_complete(); + init_hash_row_locks(&lflow_locks); + use_parallel_build = can_parallelize_hashes(false); + /* We want to detect (almost) all changes to the ovn-nb db. */ struct ovsdb_idl_loop ovnnb_idl_loop = OVSDB_IDL_LOOP_INITIALIZER( ovsdb_idl_create(ovnnb_db, &nbrec_idl_class, true, true)); @@ -13940,6 +14158,7 @@ main(int argc, char *argv[]) exiting = false; state.had_lock = false; state.paused = false; + while (!exiting) { memory_run(); if (memory_should_report()) {