From 1476b8a2d5b8778caa48da02f0ca3f5c87e86caa Mon Sep 17 00:00:00 2001 From: Se-L Date: Tue, 26 May 2020 23:21:26 +0800 Subject: [PATCH 1/2] cherry pick #3354 to release-3.1 Signed-off-by: ti-srebot --- benchmark/benchmark-tidb-using-tpcc.md | 52 +++++++++++++++++++------- 1 file changed, 39 insertions(+), 13 deletions(-) diff --git a/benchmark/benchmark-tidb-using-tpcc.md b/benchmark/benchmark-tidb-using-tpcc.md index 1c71ffde6341..bbe7d70f5ea3 100644 --- a/benchmark/benchmark-tidb-using-tpcc.md +++ b/benchmark/benchmark-tidb-using-tpcc.md @@ -57,7 +57,7 @@ ant ## 部署 TiDB 集群 -对于 1000 WAREHOUSE 我们将在 3 台服务器上部署集群。 +对于 1000 WAREHOUSE,在 3 台服务器上部署集群。 在 3 台服务器的条件下,建议每台机器部署 1 个 TiDB,1 个 PD 和 1 个 TiKV 实例。 @@ -70,23 +70,49 @@ ant | RAM | 128GB | | DISK | Optane 500GB SSD | -因为该型号 CPU 是 NUMA 架构,建议先用 `taskset` 进行绑核,首先用 `lscpu` 查看 NUMA node,比如: +因为该型号 CPU 是 NUMA 架构,建议用 `numactl` 进行绑核。 -```text -NUMA node0 CPU(s): 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38 -NUMA node1 CPU(s): 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35,37,39 -``` +1. [安装 numactl 工具](/check-before-deployment.md#安装-numactl-工具) 。 -之后可以通过下面的命令来启动 TiDB: +2. 用 `lscpu` 查看 NUMA node,比如: -{{< copyable "shell-regular" >}} + ```text + NUMA node0 CPU(s): 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38 + NUMA node1 CPU(s): 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35,37,39 + ``` -```shell -nohup taskset -c 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38 bin/tidb-server && \ -nohup taskset -c 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35,37,39 bin/tidb-server -``` +3. 通过修改 `{tidb_deploy_path}/scripts/run_tidb.sh` 启动脚本,加入 `numactl` 来启动 TiDB: + + ```text + #!/bin/bash + set -e + + ulimit -n 1000000 + + # WARNING: This file was auto-generated. Do not edit! + # All your edit might be overwritten! + DEPLOY_DIR=/home/damon/deploy/tidb1-1 + + cd "${DEPLOY_DIR}" || exit 1 + + export TZ=Asia/Shanghai + + # 同一台机器不同的 TiDB 实例需要指定不同的 cpunodebind 以及 membind;来绑定不同的 Numa node + exec numactl --cpunodebind=0 --membind=0 bin/tidb-server \ + -P 4111 \ + --status="10191" \ + --advertise-address="172.16.4.53" \ + --path="172.16.4.10:2490" \ + --config=conf/tidb.toml \ + --log-slow-query="/home/damon/deploy/tidb1-1/log/tidb_slow_query.log" \ + --log-file="/home/damon/deploy/tidb1-1/log/tidb.log" 2>> "/home/damon/deploy/tidb1-1/log/tidb_stderr.log" + ``` + + > **注意:** + > + > 直接修改 `run_tidb.sh` 可能会被覆盖,因此在生产环境中,如有绑核需求,建议使用 TiUP 绑核。 -最后,可以选择部署一个 HAproxy 来进行多个 TiDB node 的负载均衡,推荐配置 nbproc 为 CPU 核数。 +4. 选择部署一个 HAproxy 来进行多个 TiDB node 的负载均衡,推荐配置 nbproc 为 CPU 核数。 ## 配置调整 From d0c5895a85a3ec7077f22672734ef9ee3423fb0c Mon Sep 17 00:00:00 2001 From: Keke Yi <40977455+yikeke@users.noreply.github.com> Date: Fri, 19 Jun 2020 19:13:16 +0800 Subject: [PATCH 2/2] Update benchmark/benchmark-tidb-using-tpcc.md --- benchmark/benchmark-tidb-using-tpcc.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/benchmark/benchmark-tidb-using-tpcc.md b/benchmark/benchmark-tidb-using-tpcc.md index bbe7d70f5ea3..d87885f2b35c 100644 --- a/benchmark/benchmark-tidb-using-tpcc.md +++ b/benchmark/benchmark-tidb-using-tpcc.md @@ -72,7 +72,7 @@ ant 因为该型号 CPU 是 NUMA 架构,建议用 `numactl` 进行绑核。 -1. [安装 numactl 工具](/check-before-deployment.md#安装-numactl-工具) 。 +1. [安装 numactl 工具](https://docs.pingcap.com/zh/tidb/v4.0/check-before-deployment#安装-numactl-工具) 。 2. 用 `lscpu` 查看 NUMA node,比如: