作者：庆涛。 DBA，熟悉 Oracle / MySQL / SQLServer / OceanBase , 现主要从事 OceanBase 产品和解决方案推广工作。
个人公众号：OceanBase技术闲谈
OceanBase在2021年6月1日发布了社区版后（官网：https://open.oceanbase.com/)，立即有很多网友开始下载使用。部分朋友反馈在安装部署上不是很顺利，本文主要是提供一种简易的社区版安装方法，帮助大家快速了解有关原理以及具体操作步骤。
在社区版发布之前，在OceanBase 1.0版本发布之后，OceanBase软件从来只有一个版本（一个rpm包）。不管此前你从哪个渠道获取，不管是什么形态（可执行文件、rpm包、docker镜像），它跟蚂蚁内部所使用的OceanBase都是一个代码分支（功能一样）。区别只是具体的软件版本不一样（如1.4.7，2.2.3，2.2.5，2.2.7等）。在社区版发布之后，才有了社区版和企业版区分，二者功能上的确有区别，功能对比具体可以看官网 ：https://open.oceanbase.com/。
但是OceanBase最核心的基本能力都一样，包括：多副本Paxos协议强同步、SQL引擎（兼容MySQL）、分布式事务、水平扩展。所以就部署方法而言，社区版和企业版的部署方法都是一样的。以前总结的OceanBase部署经验同样适用于社区版。
参考：（https://mp.weixin.qq.com/s/uoHUd40VzIn5M_I_SbESoA）
OceanBase部署的方法比Oracle RAC简单许多，新手觉得难的主要原因在于还不太了解OB的步骤原理。当机器资源和环境都初始化正确后，命令行环境里运行5分钟即可重新部署一个OceanBase集群。很多人喜欢自动化脚本部署，觉得省事。但弊端也很明显。如果部署成功了，不知道为什么成功；部署报错了，也不知道为什么报错。除非你研究自动化部署脚本的程序 。
下面我们将提供一个方法帮助初学者把社区版快速运行起来，会省略一些OS内核参数、数据库内核参数的配置过程。因此这个方法装好的OB不适合立即展开性能测试。
首先是下载社区版。
这里不考虑源码编译安装了。按官网方法下载。
yum install -y yum-utils yum-config-manager --add-repo https://mirrors.aliyun.com/oceanbase/OceanBase.repo # 创建一个目录用于下载 mkdir rpm cat > rpm_list <<EOF oceanbase-ce-3.1.0-1.el7.x86_64.rpm oceanbase-ce-libs-3.1.0-1.el7.x86_64.rpm obproxy-3.1.0-1.el7.x86_64.rpm EOF wget -B https://mirrors.aliyun.com/oceanbase/community/stable/el/7/x86_64/ -i rpm_list -P rpm
先不要安装rpm包。
第二机器配置。
OceanBase机器里关键资源是内存。就目前而言，4C8G的环境将OB跑起来是没问题的。如果条件允许，尽可能用更大资源的机器。比如说我用的16C64G的阿里云ECS。机器资源多一点，我可以多起几个OB进程，在单机上模拟搭建一个集群。
机器磁盘也单独说一下。这关系到OB的数据文件、日志文件的存储。这个路径比较多确实有点难记。初学时简单点，机器搞一块独立的裸盘即可（/dev/sdb)。磁盘的空间稍微大于内存的4倍吧。
第三是操作系统配置。
为缩减篇幅，不说OS内核参数设置了，那些是关乎性能的。这里目标降低为跑起来。
将物理盘格式化为文件系统 /data
[root@obpilot ~]# fdisk /dev/vdb [root@obpilot ~]# mkfs -t ext4 /dev/vdb1 [root@obpilot ~]# mkdir -p /data [root@obpilot ~]# mount -t ext4 /dev/vdb1 /data
官网OBD安装的OB会运行在root用户下。这个习惯很不好。生产的OB默认运行在admin用户下，这里也这么做。要换成其他用户也可以，先按admin用户安装好熟悉一下特点再考虑换其他用户。
[root@obpilot ~]# useradd admin [root@obpilot ~]# rpm -ivh rpm/* 准备中... ################################# [100%] 正在升级/安装... 1:oceanbase-ce-libs-3.1.0-1.el7 ################################# [ 33%] 2:oceanbase-ce-3.1.0-1.el7 ################################# [ 67%] 3:obproxy-3.1.0-1.el7 ################################# [100%]
查看一下具体安装目录。
[root@obpilot ~]# rpm -ql oceanbase-ce-3.1.0-1.el7.x86_64 /home/admin/oceanbase /home/admin/oceanbase/bin /home/admin/oceanbase/bin/import_time_zone_info.py /home/admin/oceanbase/bin/observer /home/admin/oceanbase/etc /home/admin/oceanbase/etc/timezone_V1.log [root@obpilot ~]# rpm -ql obproxy-3.1.0-1.el7 /home/admin/obproxy-3.1.0/bin /home/admin/obproxy-3.1.0/bin/obproxy /home/admin/obproxy-3.1.0/bin/obproxyd.sh [root@obpilot ~]#
第四启动OB进程。
在 admin用户下启动，在oceanbase主目录启动，严格按照下面步骤执行命令。
[admin@obpilot oceanbase]$ su - admin
[admin@obpilot oceanbase]$ mkdir -p /data/observer01/store/{sort_dir,sstable,clog,ilog,slog}
[admin@obpilot oceanbase]$ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/admin/oceanbase/lib/
[admin@obpilot oceanbase]$ cd /data/observer01 && /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone1 -p 2881 -P 2882 -c 1 -d /data/observer01/store -i eth0 -l INFO
上面黄色底色的都是要根据实际环境修改的
1. -d 是指定存放 OB节点的数据文件和日志文件的目录。这个目录下必须存在几个子目录。如果你重装了，把 -d 的目录清空重新创建。
2. -r 是指定OB节点的信息。上面是单节点。后面再看三节点的示例。
3. -o 是指定OB进程启动参数，如果你熟悉OB参数，可以在这里调整。这里的 memory_limit=8G,表示启动一个8G的OB进程。如果你机器实际内存比这个大，可以改大它。内存越大的话，你甚至可以去掉这些内存相关的参数。data_size是数据文件大小，确保可以分配出来并且分配后剩余空间比例不低于10%以及是内存 2 倍以上。
4. -i 是网卡名，跟 -r 后面的IP要对应上。
5. 启动目录改到 /data/observer01下，这样相关的运行日志会在这个目录下。后面还要启动多个节点，每个节点的observer进程运行日志是分开存放的。
启动后能看到observer进程，监听端口（2881和2882），那这一步就成功了。
rs list: 172.20.249.39:2882:2881 optstr: __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 zone: zone1 mysql port: 2881 rpc port: 2882 cluster id: 1 data_dir: /data/observer01/store devname: eth0 log level: INFO [2021-06-26 17:27:34.052449] ERROR [LIB] pidfile_test (utility.cpp:1153) [9005][0][Y0-0000000000000000] [lt=0] fid file doesn't exist(pidfile="run/observer.pid") BACKTRACE:0x90a107e 0x90008fb 0x24c152f 0x251bb2d 0x90a6215 0x24be1a8 0x7f1a0a400495 0x24bd4e9 [admin@obpilot oceanbase]$ ps -ef|grep observer admin 9006 1 73 17:27 ? 00:00:04 bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone1 -p 2881 -P 2882 -c 1 -d /data/observer01/store -i eth0 -l INFO admin 9577 8043 0 17:27 pts/2 00:00:00 grep --color=auto observer [admin@obpilot oceanbase]$ netstat -ntlp (Not all processes could be identified, non-owned process infowill not be shown, you would have to be root to see it all.) Active Internet connections (only servers) Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN - tcp 0 0 0.0.0.0:2881 0.0.0.0:* LISTEN 9006/bin/observer tcp 0 0 0.0.0.0:2882 0.0.0.0:* LISTEN 9006/bin/observer [admin@obpilot oceanbase]$
说明：
如果你不是第一次启动，再次启动的时候会看到提示“ERROR [LIB] pidfile_test (utility.cpp:1153) [9005][0][Y0-0000000000000000] [lt=0] fid file doesn't exist(pidfile="run/observer.pid")”。这个报错没有实质性影响，这个信息日志级别应该用 WARN 而不是 ERROR 。希望社区版能有人把这个错误给修改了。
仅仅是进程启动了，并不代表OB部署成功了。
第五步，集群bootstrap。
不用奇怪我说集群。三副本是集群，单副本也是集群。单副本不一定是单节点。
进程第一次启动后，用mysql客户端登录（mysql版本5.5/5.6/5.7) ,root密码为空。用 obclient 也可以（需要安装 yum -y install obclient)。
[admin@obpilot oceanbase]$ mysql -h127.1 -uroot -P2881 -p Enter password: Welcome to the MariaDB monitor. Commands end with ; or \g. Your MySQL connection id is 3221225473 Server version: 5.7.25 OceanBase 3.1.0 (r-) (Built May 30 2021 11:21:29) Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. MySQL [(none)]> MySQL [(none)]>set session ob_query_timeout=1000000000; alter system bootstrap ZONE 'zone1' SERVER '172.20.249.39:2882' ; Query OK, 0 rows affected (0.00 sec) Query OK, 0 rows affected (15.67 sec) MySQL [(none)]> alter user root identified by '123456'; Query OK, 0 rows affected (0.02 sec) MySQL [(none)]> Bye [admin@obpilot oceanbase]$ mysql -h127.1 -uroot@sys -P2881 -p123456 -c -A oceanbase Welcome to the MariaDB monitor. Commands end with ; or \g. Your MySQL connection id is 3221487636 Server version: 5.7.25 OceanBase 3.1.0 (r-) (Built May 30 2021 11:21:29) Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. MySQL [oceanbase]> show databases; +--------------------+ | Database | +--------------------+ | oceanbase | | information_schema | | mysql | | SYS | | LBACSYS | | ORAAUDITOR | | test | +--------------------+ 7 rows in set (0.00 sec) MySQL [oceanbase]>
bootstrap 是学OB的第一道门槛。迈不过去就永远部署不了OB。bootstrap 原理比较复杂，根据原理或日志去找原因不是个好主意。bootstrap失败的原因通常很简单。如内存资源不够、目录权限不对、磁盘空间不够等等、多节点时间不同步（超过5ms）、多节点网络延时很大（超过50ms）。这一块有单独的文章总结，可以参考《OceanBase 2.2安装部署问题解答》。
bootstrap 成功后，用用户名 root@sys 登录后，查看数据库，可以看到多出几个database：oceanbase 。
看到 oceanbase 这个数据库，那就是真正成功了。
第六步，建租户（实例）。
租户就是实例，是逻辑实例，没有单独的进程。OceanBase在企业自己机房部署后，可以给开发提供类似云数据库的服务。Database As A Service。实例按需分配，立即分配（1分钟不到），实例资源后期可以在线调整，多退少补。OceanBase改变的是传统数据库的运维形态，DBA更像是一个资源管家，更专注于服务业务。
这一步很简单，但是很多人也会在这里碰到资源不足问题。在分配租户之前，首先得知道集群有多少资源可用。
MySQL [oceanbase]> select a.zone,concat(a.svr_ip,':',a.svr_port) observer, cpu_total, (cpu_total-cpu_assigned) cpu_free, round(mem_total/1024/1024/1024) mem_total_gb, round((mem_total-mem_assigned)/1024/1024/1024) mem_free_gb, usec_to_time(b.last_offline_time) last_offline_time, usec_to_time(b.start_service_time) start_service_time, b.status, b.build_version , usec_to_time(a.stop_time) stop_time
from __all_virtual_server_stat a join __all_server b on (a.svr_ip=b.svr_ip and a.svr_port=b.svr_port)
order by a.zone, a.svr_ip
;
+-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+
| zone | observer | cpu_total | cpu_free | mem_total_gb | mem_free_gb | last_offline_time | start_service_time | status | build_version | stop_time |
+-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+
| zone1 | 172.20.249.39:2882 | 14 | 11.5 | 4 | 3 | 1970-01-01 08:00:00.000000 | 2021-06-26 17:29:25.599983 | active | 3.1.0_-(May 30 2021 11:21:29) | 1970-01-01 08:00:00.000000 |
+-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+
1 row in set (0.00 sec)
MySQL [oceanbase]> select t1.name resource_pool_name, t2.`name` unit_config_name, t2.max_cpu, t2.min_cpu, round(t2.max_memory/1024/1024/1024) max_mem_gb, round(t2.min_memory/1024/1024/1024) min_mem_gb, t3.unit_id, t3.zone, concat(t3.svr_ip,':',t3.`svr_port`) observer,t4.tenant_id, t4.tenant_name
from __all_resource_pool t1 join __all_unit_config t2 on (t1.unit_config_id=t2.unit_config_id)
join __all_unit t3 on (t1.`resource_pool_id` = t3.`resource_pool_id`)
left join __all_tenant t4 on (t1.tenant_id=t4.tenant_id)
order by t1.`resource_pool_id`, t2.`unit_config_id`, t3.unit_id
;
-> ;
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
| resource_pool_name | unit_config_name | max_cpu | min_cpu | max_mem_gb | min_mem_gb | unit_id | zone | observer | tenant_id | tenant_name |
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
| sys_pool | sys_unit_config | 5 | 2.5 | 1 | 1 | 1 | zone1 | 172.20.249.39:2882 | 1 | sys |
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
1 row in set (0.00 sec)
第一个sql 是查集群各个节点的剩余资源（cpu和内存，空间不管）。第二个sql是查询集群里当前资源使用细节。这里有几点解释：
1. observer 启动参数里指定了cpu_count=16，表示告诉（欺骗）observer进程主机有16个cpu，实际上observer进程只拿了14个cpu，留了2个给os。生产环境实打实的。
2. observer 启动进程里指定了memory_limit=8G，表示告诉（不是欺骗）observer进程可用8G内存，ob内部system_memory拿去了4G，还剩下4G.所以这里的memory_total是4G，然后内部租户 sys_pool用了1G，留给租户的就是3G。
细心的朋友如果计算一下cpu的分配，总cpu是14个，sys租户用了2.5，剩余是11.5.实际上实际可用的cpu不是11.5，而是9个。这个是因为 sys_unit_config的min_cpu值跟max_cpu不一样（在企业版本里，min_memory跟max_memory也不一样）。而很多人根据第一个sql计算觉得资源够，实际分配的时候不够往往就是这个原因。
所以，个人习惯，首先把这个不协调的设置改对。真希望社区版能有人把这个代码逻辑改过来。
MySQL [oceanbase]> alter resource unit sys_unit_config min_cpu=5,min_memory='1G'; Query OK, 0 rows affected (0.01 sec) MySQL [oceanbase]> select a.zone,concat(a.svr_ip,':',a.svr_port) observer, cpu_total, (cpu_total-cpu_assigned) cpu_free, round(mem_total/1024/1024/1024) mem_total_gb, round((mem_total-mem_assigned)/1024/1024/1024) mem_free_gb, usec_to_time(b.last_offline_time) last_offline_time, usec_to_time(b.start_service_time) start_service_time, b.status, b.build_version , usec_to_time(a.stop_time) stop_time from __all_virtual_server_stat a join __all_server b on (a.svr_ip=b.svr_ip and a.svr_port=b.svr_port) order by a.zone, a.svr_ip; +-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+ | zone | observer | cpu_total | cpu_free | mem_total_gb | mem_free_gb | last_offline_time | start_service_time | status | build_version | stop_time | +-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+ | zone1 | 172.20.249.39:2882 | 14 | 9 | 4 | 3 | 1970-01-01 08:00:00.000000 | 2021-06-26 17:29:25.599983 | active | 3.1.0_-(May 30 2021 11:21:29) | 1970-01-01 08:00:00.000000 | +-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+ 1 row in set (0.01 sec)
现在可以愉快的创建租户（实例）了。
MySQL [oceanbase]> CREATE resource unit my_unit_config max_cpu=4, min_cpu=4, max_memory='1G', min_memory='1G', max_iops=10000, min_iops=1000, max_session_num=1000000, max_disk_size='1024G';
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> CREATE resource pool pool_mysql_01 unit = 'my_unit_config', unit_num = 1;
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> CREATE tenant mysql01 resource_pool_list=('pool_mysql_01'), primary_zone='RANDOM',comment 'mysql tenant/instance', charset='utf8' set ob_tcp_invited_nodes='%', ob_compatibility_mode='mysql';
Query OK, 0 rows affected (0.48 sec)
MySQL [oceanbase]>
MySQL [oceanbase]> CREATE resource pool pool_mysql_02 unit = 'my_unit_config', unit_num = 1;
Query OK, 0 rows affected (0.00 sec)
MySQL [oceanbase]>
MySQL [oceanbase]> CREATE tenant mysql02 resource_pool_list=('pool_mysql_02'), primary_zone='RANDOM',comment 'mysql tenant/instance', charset='utf8' set ob_tcp_invited_nodes='%', ob_compatibility_mode='mysql';
Query OK, 0 rows affected (0.46 sec)
我这里创建了2个租户。4C1G的mysql租户，就看看功能就行。不适合大量建表和导入数据，以及性能测试。
MySQL [oceanbase]> select * from __all_tenant;
+----------------------------+----------------------------+-----------+-------------+-------------+-----------+--------------+--------+----------------+-----------------------+-----------+-----------------------+---------------+---------------------+-------------------+------------------------+-----------------------------+-----------------------+--------------------+------------------+----------------------+---------------+
| gmt_create | gmt_modified | tenant_id | tenant_name | replica_num | zone_list | primary_zone | locked | collation_type | info | read_only | rewrite_merge_version | locality | logonly_replica_num | previous_locality | storage_format_version | storage_format_work_version | default_tablegroup_id | compatibility_mode | drop_tenant_time | status | in_recyclebin |
+----------------------------+----------------------------+-----------+-------------+-------------+-----------+--------------+--------+----------------+-----------------------+-----------+-----------------------+---------------+---------------------+-------------------+------------------------+-----------------------------+-----------------------+--------------------+------------------+----------------------+---------------+
| 2021-06-26 17:29:24.242717 | 2021-06-26 17:29:24.242717 | 1 | sys | -1 | zone1 | zone1 | 0 | 0 | system tenant | 0 | 0 | FULL{1}@zone1 | 0 | | 0 | 0 | -1 | 0 | -1 | TENANT_STATUS_NORMAL | 0 |
| 2021-06-26 19:20:13.030153 | 2021-06-26 19:20:13.030153 | 1001 | mysql01 | -1 | zone1 | RANDOM | 0 | 0 | mysql tenant/instance | 0 | 0 | FULL{1}@zone1 | 0 | | 0 | 0 | -1 | 0 | -1 | TENANT_STATUS_NORMAL | 0 |
| 2021-06-26 19:20:58.340233 | 2021-06-26 19:20:58.340233 | 1003 | mysql02 | -1 | zone1 | RANDOM | 0 | 0 | mysql tenant/instance | 0 | 0 | FULL{1}@zone1 | 0 | | 0 | 0 | -1 | 0 | -1 | TENANT_STATUS_NORMAL | 0 |
+----------------------------+----------------------------+-----------+-------------+-------------+-----------+--------------+--------+----------------+-----------------------+-----------+-----------------------+---------------+---------------------+-------------------+------------------------+-----------------------------+-----------------------+--------------------+------------------+----------------------+---------------+
3 rows in set (0.00 sec)
select t1.name resource_pool_name, t2.`name` unit_config_name, t2.max_cpu, t2.min_cpu, round(t2.max_memory/1024/1024/1024) max_mem_gb, round(t2.min_memory/1024/1024/1024) min_mem_gb, t3.unit_id, t3.zone, concat(t3.svr_ip,':',t3.`svr_port`) observer,t4.tenant_id, t4.tenant_name
from __all_resource_pool t1 join __all_unit_config t2 on (t1.unit_config_id=t2.unit_config_id)
join __all_unit t3 on (t1.`resource_pool_id` = t3.`resource_pool_id`)
left join __all_tenant t4 on (t1.tenant_id=t4.tenant_id)
order by t1.`resource_pool_id`, t2.`unit_config_id`, t3.unit_id
;
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
| resource_pool_name | unit_config_name | max_cpu | min_cpu | max_mem_gb | min_mem_gb | unit_id | zone | observer | tenant_id | tenant_name |
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
| sys_pool | sys_unit_config | 5 | 5 | 1 | 1 | 1 | zone1 | 172.20.249.39:2882 | 1 | sys |
| pool_mysql_01 | my_unit_config | 4 | 4 | 1 | 1 | 1001 | zone1 | 172.20.249.39:2882 | 1001 | mysql01 |
| pool_mysql_02 | my_unit_config | 4 | 4 | 1 | 1 | 1002 | zone1 | 172.20.249.39:2882 | 1003 | mysql02 |
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
3 rows in set (0.00 sec)
在往后就是OB的使用了。还有obproxy的部署、租户的连接。这些跟企业版也是一样的。有问题的朋友可以查看以前的文章。这里不再赘述了。
特别介绍：单副本扩容到三副本。
这里再介绍一下从单节点扩容到三节点，能看懂了表示就理解了OceanBase集群扩容原理。
chown -R admin.admin /data
su - admin
echo "export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/admin/oceanbase/lib/" >> ~/.bash_profile
. ~/.bash_profile
mkdir -p /data/observer02/store/{sort_dir,sstable,clog,ilog,slog}
cd /data/observer02/ && /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone2 -p 3881 -P 3882 -c 1 -d /data/observer02/store -i eth0 -l INFO
su - admin
mkdir -p /data/observer03/store/{sort_dir,sstable,clog,ilog,slog}
cd /data/observer03/ && /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone3 -p 4881 -P 4882 -c 1 -d /data/observer03/store -i eth0 -l INFO
[admin@obpilot observer03]$ ps -ef|grep observer |grep -v grep
admin 26533 1 50 19:46 ? 00:02:57 /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone1 -p 2881 -P 2882 -c 1 -d /data/observer01/store -i eth0 -l INFO
admin 27316 1 59 19:50 ? 00:01:01 /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone2 -p 3881 -P 3882 -c 1 -d /data/observer02/store -i eth0 -l INFO
admin 27955 1 72 19:52 ? 00:00:11 /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone3 -p 4881 -P 4882 -c 1 -d /data/observer03/store -i eth0 -l INFO
[admin@obpilot observer03]$ netstat -ntlp
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
tcp 0 0 0.0.0.0:3881 0.0.0.0:* LISTEN 27316/observer
tcp 0 0 0.0.0.0:3882 0.0.0.0:* LISTEN 27316/observer
tcp 0 0 0.0.0.0:4881 0.0.0.0:* LISTEN 27955/observer
tcp 0 0 0.0.0.0:4882 0.0.0.0:* LISTEN 27955/observer
tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN -
tcp 0 0 0.0.0.0:2881 0.0.0.0:* LISTEN 26533/observer
tcp 0 0 0.0.0.0:2882 0.0.0.0:* LISTEN 26533/observer
[admin@obpilot observer03]
说明：
1. -r 参数依然是第一个节点的端口（2881和2882），因为这里是集群扩容，不是新增2个集群。所以 -r 参数会决定了这个进程启动后跟哪个集群进行通信。
2. -p 和 -P 参数要不一样。每个进程监听2个端口，不能冲突。
3. 不同节点的启动目录要不一样，这样每个目录下会生成相应的log目录，里面有运行日志。
进程启动了，还要在集群里把这两个节点加进集群。
MySQL [oceanbase]> select * from __all_zone where name in ('region','status','zone_type');
+----------------------------+----------------------------+-------+-----------+-------+----------------+
| gmt_create | gmt_modified | zone | name | value | info |
+----------------------------+----------------------------+-------+-----------+-------+----------------+
| 2021-06-26 20:01:31.466919 | 2021-06-26 20:01:31.466919 | zone1 | region | 0 | default_region |
| 2021-06-26 20:01:31.465863 | 2021-06-26 20:01:31.465863 | zone1 | status | 2 | ACTIVE |
| 2021-06-26 20:01:31.466919 | 2021-06-26 20:01:31.466919 | zone1 | zone_type | 0 | ReadWrite |
+----------------------------+----------------------------+-------+-----------+-------+----------------+
3 rows in set (0.00 sec)
MySQL [oceanbase]> alter system add zone 'zone2' region 'default_region';
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> alter system add zone 'zone3' region 'default_region';
Query OK, 0 rows affected (0.00 sec)
MySQL [oceanbase]> alter system start zone 'zone2';
Query OK, 0 rows affected (0.00 sec)
MySQL [oceanbase]> alter system start zone 'zone3';
Query OK, 0 rows affected (0.00 sec)
MySQL [oceanbase]> select * from __all_zone where name in ('region','status','zone_type');
+----------------------------+----------------------------+-------+-----------+-------+----------------+
| gmt_create | gmt_modified | zone | name | value | info |
+----------------------------+----------------------------+-------+-----------+-------+----------------+
| 2021-06-26 20:01:31.466919 | 2021-06-26 20:01:31.466919 | zone1 | region | 0 | default_region |
| 2021-06-26 20:01:31.465863 | 2021-06-26 20:01:31.465863 | zone1 | status | 2 | ACTIVE |
| 2021-06-26 20:01:31.466919 | 2021-06-26 20:01:31.466919 | zone1 | zone_type | 0 | ReadWrite |
| 2021-06-26 20:07:15.488710 | 2021-06-26 20:07:15.488710 | zone2 | region | 0 | default_region |
| 2021-06-26 20:07:15.488508 | 2021-06-26 20:08:31.962846 | zone2 | status | 2 | ACTIVE |
| 2021-06-26 20:07:15.489765 | 2021-06-26 20:07:15.489765 | zone2 | zone_type | 0 | LOCAL |
| 2021-06-26 20:07:20.821128 | 2021-06-26 20:07:20.821128 | zone3 | region | 0 | default_region |
| 2021-06-26 20:07:20.820098 | 2021-06-26 20:08:34.522818 | zone3 | status | 2 | ACTIVE |
| 2021-06-26 20:07:20.821128 | 2021-06-26 20:07:20.821128 | zone3 | zone_type | 0 | LOCAL |
+----------------------------+----------------------------+-------+-----------+-------+----------------+
9 rows in set (0.00 sec)
MySQL [oceanbase]> alter system add server '172.20.249.39:3882' zone 'zone2';
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> alter system add server '172.20.249.39:4882' zone 'zone3';
Query OK, 0 rows affected (0.01 sec)
select a.zone,concat(a.svr_ip,':',a.svr_port) observer, cpu_total, (cpu_total-cpu_assigned) cpu_free, round(mem_total/1024/1024/1024) mem_total_gb, round((mem_total-mem_assigned)/1024/1024/1024) mem_free_gb, usec_to_time(b.last_offline_time) last_offline_time, usec_to_time(b.start_service_time) start_service_time, b.status, b.build_version , usec_to_time(a.stop_time) stop_time
from __all_virtual_server_stat a join __all_server b on (a.svr_ip=b.svr_ip and a.svr_port=b.svr_port)
order by a.zone, a.svr_ip
;
+-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+
| zone | observer | cpu_total | cpu_free | mem_total_gb | mem_free_gb | last_offline_time | start_service_time | status | build_version | stop_time |
+-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+
| zone1 | 172.20.249.39:2882 | 14 | 1 | 4 | 1 | 1970-01-01 08:00:00.000000 | 2021-06-26 20:01:32.824604 | active | 3.1.0_-(May 30 2021 11:21:29) | 1970-01-01 08:00:00.000000 |
| zone2 | 172.20.249.39:3882 | 14 | 14 | 4 | 4 | 1970-01-01 08:00:00.000000 | 2021-06-26 20:01:35.617217 | active | 3.1.0_-(May 30 2021 11:21:29) | 1970-01-01 08:00:00.000000 |
| zone3 | 172.20.249.39:4882 | 14 | 14 | 4 | 4 | 1970-01-01 08:00:00.000000 | 2021-06-26 20:09:21.226741 | active | 3.1.0_-(May 30 2021 11:21:29) | 1970-01-01 08:00:00.000000 |
+-------+--------------------+-----------+----------+--------------+-------------+----------------------------+----------------------------+--------+-------------------------------+----------------------------+
3 rows in set (0.00 sec)
MySQL [oceanbase]> CREATE resource pool pool_mysql_01b unit = 'my_unit_config', unit_num = 1, zone_list=('zone2','zone3');
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> CREATE resource pool pool_mysql_02b unit = 'my_unit_config', unit_num = 1, zone_list=('zone2','zone3');
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> alter tenant mysql01 resource_pool_list=('pool_mysql_01','pool_mysql_01b');
Query OK, 0 rows affected (0.03 sec)
MySQL [oceanbase]> alter tenant mysql01 locality='FULL{1}@zone1, FULL{1}@zone2, FULL{1}@zone3';
ERROR 4179 (HY000): violate locality principal not allowed
MySQL [oceanbase]> alter tenant mysql01 primary_zone='zone1';
Query OK, 0 rows affected (0.02 sec)
MySQL [oceanbase]> alter tenant mysql01 locality='FULL{1}@zone1, FULL{1}@zone2, FULL{1}@zone3';
ERROR 4179 (HY000): violate locality principal not allowed
MySQL [oceanbase]> alter tenant mysql01 locality='FULL{1}@zone1, FULL{1}@zone2';
Query OK, 0 rows affected (0.02 sec)
MySQL [oceanbase]> alter tenant mysql01 locality='FULL{1}@zone1, FULL{1}@zone2, FULL{1}@zone3';
ERROR 4179 (HY000): alter tenant locality when previous operation is in progress not allowed
MySQL [oceanbase]> select gmt_create,gmt_modified,job_id,job_type,job_status,return_code,progress,tenant_id from __all_rootservice_job;
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
| gmt_create | gmt_modified | job_id | job_type | job_status | return_code | progress | tenant_id |
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
| 2021-06-26 20:12:54.409071 | 2021-06-26 20:13:25.724652 | 1 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1001 |
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
1 row in set (0.01 sec)
MySQL [oceanbase]> CREATE resource pool sys_poolb unit = 'sys_unit_config', unit_num = 1, zone_list=('zone2','zone3');
Query OK, 0 rows affected (0.01 sec)
MySQL [oceanbase]> alter tenant sys resource_pool_list=('sys_pool','sys_poolb');
Query OK, 0 rows affected (0.02 sec)
MySQL [oceanbase]> alter tenant sys locality='FULL{1}@zone1, FULL{1}@zone2';
Query OK, 0 rows affected (0.02 sec)
MySQL [oceanbase]> alter tenant sys locality='FULL{1}@zone1, FULL{1}@zone2, FULL{1}@zone3';
ERROR 4179 (HY000): alter tenant locality when previous operation is in progress not allowed
MySQL [oceanbase]> select gmt_create,gmt_modified,job_id,job_type,job_status,return_code,progress,tenant_id from __all_rootservice_job;
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
| gmt_create | gmt_modified | job_id | job_type | job_status | return_code | progress | tenant_id |
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
| 2021-06-26 20:12:54.409071 | 2021-06-26 20:13:25.724652 | 1 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1001 |
| 2021-06-26 20:15:16.628450 | 2021-06-26 20:15:48.679600 | 2 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1001 |
| 2021-06-26 20:19:37.365538 | 2021-06-26 20:19:37.365538 | 5 | ALTER_TENANT_LOCALITY | INPROGRESS | NULL | 0 | 1 |
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
5 rows in set (0.00 sec)
-- 等若干分钟
MySQL [oceanbase]> alter tenant sys locality='FULL{1}@zone1, FULL{1}@zone2, FULL{1}@zone3';
Query OK, 0 rows affected (0.03 sec)
MySQL [oceanbase]> alter tenant mysql02 resource_pool_list=('pool_mysql_02','pool_mysql_02b');
Query OK, 0 rows affected (0.05 sec)
MySQL [oceanbase]> alter tenant mysql02 locality='FULL{1}@zone1, FULL{1}@zone2';
Query OK, 0 rows affected (0.02 sec)
-- 等若干分钟
MySQL [oceanbase]> alter tenant mysql02 locality='FULL{1}@zone1, FULL{1}@zone2, FULL{1}@zone3';
Query OK, 0 rows affected (0.01 sec)
-- 等若干分钟
MySQL [oceanbase]> select gmt_create,gmt_modified,job_id,job_type,job_status,return_code,progress,tenant_id from __all_rootservice_job;
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
| gmt_create | gmt_modified | job_id | job_type | job_status | return_code | progress | tenant_id |
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
| 2021-06-26 20:12:54.409071 | 2021-06-26 20:13:25.724652 | 1 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1001 |
| 2021-06-26 20:15:16.628450 | 2021-06-26 20:15:48.679600 | 2 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1001 |
| 2021-06-26 20:19:37.365538 | 2021-06-26 20:23:15.879598 | 5 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1 |
| 2021-06-26 20:23:24.097481 | 2021-06-26 20:23:40.449128 | 6 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1002 |
| 2021-06-26 20:23:42.709761 | 2021-06-26 20:26:18.419506 | 7 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1 |
| 2021-06-26 20:24:13.183605 | 2021-06-26 20:26:33.958669 | 8 | ALTER_TENANT_LOCALITY | SUCCESS | 0 | 100 | 1002 |
+----------------------------+----------------------------+--------+-----------------------+------------+-------------+----------+-----------+
8 rows in set (0.01 sec)
这里面信息量非常大，简单说明一下：
1. 先add zone ，要逐个加，然后是逐个 start zone 。
2. 再add server，要指定 节点监听端口和zone，这个跟节点启动进程参数要对应。如果报错，说明没对上。
3. 再给各个租户补充资源池里资源单元。create resource pool 可以指定 zone_list。然后 alter tenant resource_pool_list 指定资源池。一个租户在每个zone只能有一个资源池，但是在不同zone可以有多个资源池。
4. 修改租户的locality属性。做之前，先设置明确的primary_zone(不要random）。修改locality必须缓慢的从一个zone到两个zone，这期间后台会做数据补副本操作，根据实际数据量需要点时间。结束了才能做两个zone到三个zone的扩容。
OB的扩容命令就是这么简单。
最后检查一下扩容结果。
select t1.name resource_pool_name, t2.`name` unit_config_name, t2.max_cpu, t2.min_cpu, round(t2.max_memory/1024/1024/1024) max_mem_gb, round(t2.min_memory/1024/1024/1024) min_mem_gb, t3.unit_id, t3.zone, concat(t3.svr_ip,':',t3.`svr_port`) observer,t4.tenant_id, t4.tenant_name
from __all_resource_pool t1 join __all_unit_config t2 on (t1.unit_config_id=t2.unit_config_id)
join __all_unit t3 on (t1.`resource_pool_id` = t3.`resource_pool_id`)
left join __all_tenant t4 on (t1.tenant_id=t4.tenant_id)
order by t4.tenant_id, t1.`resource_pool_id`, t2.`unit_config_id`, t3.unit_id
;
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
| resource_pool_name | unit_config_name | max_cpu | min_cpu | max_mem_gb | min_mem_gb | unit_id | zone | observer | tenant_id | tenant_name |
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
| sys_pool | sys_unit_config | 5 | 5 | 1 | 1 | 1 | zone1 | 172.20.249.39:2882 | 1 | sys |
| sys_poolb | sys_unit_config | 5 | 5 | 1 | 1 | 1007 | zone2 | 172.20.249.39:3882 | 1 | sys |
| sys_poolb | sys_unit_config | 5 | 5 | 1 | 1 | 1008 | zone3 | 172.20.249.39:4882 | 1 | sys |
| pool_mysql_01 | my_unit_config | 4 | 4 | 1 | 1 | 1001 | zone1 | 172.20.249.39:2882 | 1001 | mysql01 |
| pool_mysql_01b | my_unit_config | 4 | 4 | 1 | 1 | 1003 | zone2 | 172.20.249.39:3882 | 1001 | mysql01 |
| pool_mysql_01b | my_unit_config | 4 | 4 | 1 | 1 | 1004 | zone3 | 172.20.249.39:4882 | 1001 | mysql01 |
| pool_mysql_02 | my_unit_config | 4 | 4 | 1 | 1 | 1002 | zone1 | 172.20.249.39:2882 | 1002 | mysql02 |
| pool_mysql_02b | my_unit_config | 4 | 4 | 1 | 1 | 1005 | zone2 | 172.20.249.39:3882 | 1002 | mysql02 |
| pool_mysql_02b | my_unit_config | 4 | 4 | 1 | 1 | 1006 | zone3 | 172.20.249.39:4882 | 1002 | mysql02 |
+--------------------+------------------+---------+---------+------------+------------+---------+-------+--------------------+-----------+-------------+
9 rows in set (0.00 sec)
特别介绍：obproxy部署。
obproxy的部署方法跟早期文章里写的方法有点变化了。
首先，obproxy跟OB集群通信是使用sys租户内的一个内部账户proxyro。这个账户需要创建。
[admin@obpilot obproxy]$ mysql -h127.1 -uroot@sys -P2881 -p123456 -c -A oceanbase MySQL [oceanbase]> create user proxyro identified by 'proxyro123456'; Query OK, 0 rows affected (0.01 sec) MySQL [oceanbase]> grant select on oceanbase.* to proxyro; Query OK, 0 rows affected (0.02 sec)
然后开始启动obproxy。
mkdir -p /data/obproxy cd /data/obproxy && /home/admin/obproxy-3.1.0/bin/obproxy -p2883 -c obcluster -r "172.20.249.39:2881;172.20.249.39:3881;172.20.249.39:4881" -o "enable_cluster_checkout=false,enable_strict_kernel_release=false,enable_metadb_used=false" mysql -h127.1 -uroot@proxysys -P2883 -p MySQL [(none)]> alter proxyconfig set obproxy_sys_password='proxysys123456'; Query OK, 0 rows affected (0.00 sec) MySQL [(none)]> alter proxyconfig set observer_sys_password='proxyro123456'; Query OK, 0 rows affected (0.00 sec) MySQL [(none)]> show proxyconfig like '%sys_password%'; +-----------------------+------------------------------------------+--------------------------------+-------------+---------------+ | name | value | info | need_reboot | visible_level | +-----------------------+------------------------------------------+--------------------------------+-------------+---------------+ | observer_sys_password | dcb332deaee8b0030c3fd21a8e9f2813984339ea | password for observer sys user | false | SYS | | obproxy_sys_password | 3a8d187554571d6be2b27749187dcb8ca6285ede | password for obproxy sys user | false | SYS | +-----------------------+------------------------------------------+--------------------------------+-------------+---------------+ 2 rows in set (0.00 sec) [admin@obpilot obproxy]$ mysql -h127.1 -uroot@sys#obcluster -P2883 -p123456 -c -A oceanbase MySQL [oceanbase]> show full processlist; +------------+---------+--------+---------------------+-----------+---------+------+--------+-----------------------+---------------+------+--------------+ | Id | User | Tenant | Host | db | Command | Time | State | Info | Ip | Port | Proxy_sessid | +------------+---------+--------+---------------------+-----------+---------+------+--------+-----------------------+---------------+------+--------------+ | 3221633341 | root | sys | 172.20.249.39:41650 | oceanbase | Query | 0 | ACTIVE | show full processlist | 172.20.249.39 | 2881 | 2 | | 3221749764 | proxyro | sys | 172.20.249.39:32942 | oceanbase | Sleep | 6 | SLEEP | NULL | 172.20.249.39 | 3881 | 3 | +------------+---------+--------+---------------------+-----------+---------+------+--------+-----------------------+---------------+------+--------------+ 2 rows in set (0.01 sec)
说明：
1. -r 是指定OB集群的rootservice_list 地址，因为OB已经扩容为三副本里，所以这里会有三个地址（ip:port,port使用连接端口）。
2. 给obproxy指定一个启动目录 /data/obproxy。这个可以改。如果启动多个obproxy进程，就设置多个目录。同样，obproxy运行在admin用户下。多个obproxy就要设置不同的监听端口。默认是2883.
3. obproxy启动后，默认用 root@proxysys 登录，密码为空。需要改密码（通过proxy参数obproxy_sys_password指定）。
4. obproxy启动后，还需要修改proxyro的密码（通过proxy参数observer_sys_password指定），设置为跟OB集群里创建的proxyro密码一致才能链接那个OB集群。
测试一下这个obproxy。
[admin@obpilot obproxy]$ mysql -h127.1 -uroot@sys#obcluster -P2883 -p123456 -c -A oceanbase Welcome to the MariaDB monitor. Commands end with ; or \g. Your MySQL connection id is 5 Server version: 5.6.25 OceanBase 3.1.0 (r-) (Built May 30 2021 11:21:29) Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. MySQL [oceanbase]> select * from __all_server; +----------------------------+----------------------------+---------------+----------+----+-------+------------+-----------------+--------+-----------------------+-------------------------------+-----------+--------------------+--------------+----------------+-------------------+ | gmt_create | gmt_modified | svr_ip | svr_port | id | zone | inner_port | with_rootserver | status | block_migrate_in_time | build_version | stop_time | start_service_time | first_sessid | with_partition | last_offline_time | +----------------------------+----------------------------+---------------+----------+----+-------+------------+-----------------+--------+-----------------------+-------------------------------+-----------+--------------------+--------------+----------------+-------------------+ | 2021-06-26 20:01:24.753198 | 2021-06-26 20:01:33.821660 | 172.20.249.39 | 2882 | 1 | zone1 | 2881 | 1 | active | 0 | 3.1.0_-(May 30 2021 11:21:29) | 0 | 1624708892824604 | 0 | 1 | 0 | | 2021-06-26 20:09:02.347295 | 2021-06-26 20:12:54.900390 | 172.20.249.39 | 3882 | 2 | zone2 | 3881 | 0 | active | 0 | 3.1.0_-(May 30 2021 11:21:29) | 0 | 1624708895617217 | 0 | 1 | 0 | | 2021-06-26 20:09:10.246956 | 2021-06-26 20:15:17.112561 | 172.20.249.39 | 4882 | 3 | zone3 | 4881 | 0 | active | 0 | 3.1.0_-(May 30 2021 11:21:29) | 0 | 1624709361226741 | 0 | 1 | 0 | +----------------------------+----------------------------+---------------+----------+----+-------+------------+-----------------+--------+-----------------------+-------------------------------+-----------+--------------------+--------------+----------------+-------------------+ 3 rows in set (0.02 sec) MySQL [oceanbase]> show full processlist; +------------+---------+--------+---------------------+-----------+---------+------+--------+-----------------------+---------------+------+--------------+ | Id | User | Tenant | Host | db | Command | Time | State | Info | Ip | Port | Proxy_sessid | +------------+---------+--------+---------------------+-----------+---------+------+--------+-----------------------+---------------+------+--------------+ | 3222011908 | root | sys | 172.20.249.39:60006 | oceanbase | Sleep | 19 | SLEEP | NULL | 172.20.249.39 | 4881 | 4 | | 3221749764 | proxyro | sys | 172.20.249.39:32942 | oceanbase | Sleep | 6 | SLEEP | NULL | 172.20.249.39 | 3881 | 3 | | 3221649264 | root | sys | 172.20.249.39:41704 | oceanbase | Query | 0 | ACTIVE | show full processlist | 172.20.249.39 | 2881 | 4 | +------------+---------+--------+---------------------+-----------+---------+------+--------+-----------------------+---------------+------+--------------+ 3 rows in set (0.01 sec) MySQL [oceanbase]>
至此，obproxy安装成功。
特别介绍：OB三副本直接部署。
上面可能会有人误解为OB安装需要从单副本开始，再扩容到三副本。这里再提供一下三副本的直接安装方法。
首先彻底清理掉已安装的OB和OBPROXY。
kill -9 `pidof observer` `pidof obproxy`
sleep 3
/bin/rm -rf /data/observer{01,02,03} /data/obproxy
开始启动三个OB节点进程。
mkdir -p /data/{observer01,observer02,observer03,obproxy}
mkdir -p /data/observer{01,02,03}/store/{sort_dir,sstable,clog,ilog,slog}
cd /data/observer01/ && /home/admin/oceanbase/bin/observer -r "172.20.249.39:2882:2881;172.20.249.39:3882:3881;172.20.249.39:4882:4881" -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone1 -p 2881 -P 2882 -c 1 -d /data/observer01/store -i eth0 -l INFO
cd /data/observer02/ && /home/admin/oceanbase/bin/observer -r "172.20.249.39:2882:2881;172.20.249.39:3882:3881;172.20.249.39:4882:4881" -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone2 -p 3881 -P 3882 -c 1 -d /data/observer02/store -i eth0 -l INFO
cd /data/observer03/ && /home/admin/oceanbase/bin/observer -r "172.20.249.39:2882:2881;172.20.249.39:3882:3881;172.20.249.39:4882:4881" -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone3 -p 4881 -P 4882 -c 1 -d /data/observer03/store -i eth0 -l INFO
一样的，检查三个进程都监听成功。
[admin@obpilot observer03]$ ps -ef|grep observer |grep -v grep admin 21194 1 89 13:51 ? 00:04:20 /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881;172.20.249.39:3882:3881;172.20.249.39:4882:4881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone1 -p 2881 -P 2882 -c 1 -d /data/observer01/store -i eth0 -l INFO admin 21775 1 82 13:52 ? 00:03:47 /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881;172.20.249.39:3882:3881;172.20.249.39:4882:4881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone2 -p 3881 -P 3882 -c 1 -d /data/observer02/store -i eth0 -l INFO admin 22371 1 86 13:52 ? 00:03:32 /home/admin/oceanbase/bin/observer -r 172.20.249.39:2882:2881;172.20.249.39:3882:3881;172.20.249.39:4882:4881 -o __min_full_resource_pool_memory=268435456,memory_limit=8G,system_memory=4G,stack_size=512K,cpu_count=16,cache_wash_threshold=1G,workers_per_cpu_quota=10,schema_history_expire_time=1d,net_thread_count=4,sys_bkgd_migration_retry_num=3,minor_freeze_times=10,enable_separate_sys_clog=0,enable_merge_by_turn=False,datafile_size=50G,enable_syslog_recycle=True,max_syslog_file_count=10 -z zone3 -p 4881 -P 4882 -c 1 -d /data/observer03/store -i eth0 -l INFO [admin@obpilot observer03]$ netstat -ntlp (Not all processes could be identified, non-owned process info will not be shown, you would have to be root to see it all.) Active Internet connections (only servers) Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name tcp 0 0 0.0.0.0:3881 0.0.0.0:* LISTEN 21775/observer tcp 0 0 0.0.0.0:3882 0.0.0.0:* LISTEN 21775/observer tcp 0 0 0.0.0.0:4881 0.0.0.0:* LISTEN 22371/observer tcp 0 0 0.0.0.0:4882 0.0.0.0:* LISTEN 22371/observer tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN - tcp 0 0 0.0.0.0:2881 0.0.0.0:* LISTEN 21194/observer tcp 0 0 0.0.0.0:2882 0.0.0.0:* LISTEN 21194/observer [admin@obpilot observer03]$
修改几个账户密码
mysql -h172.20.249.39 -uroot -P2881 -p -c -A set session ob_query_timeout=1000000000; alter system bootstrap ZONE 'zone1' SERVER '172.20.249.39:2882' , ZONE 'zone2' SERVER '172.20.249.39:3882' , ZONE 'zone3' SERVER '172.20.249.39:4882'; alter user root identified by '123456'; create user proxyro identified by 'proxyro123456';
安装obproxy的方法同上。obproxy启动时指定 rootservice_list写死这个地址不是很灵活。在企业版里，OB会把rootservice_list地址通过OCP API保存到元数据库里。
最后说一些常见问题解答。
- observer.log 比较难懂，找你看得懂的ERROR日志。或者根据错误号到官网查看说明。总会找到有用的线索。看得懂源码的朋友例外。
- 部署是一定能成功的，不成功就是某一个细节没做好。看上面步骤细节并不多。此外机器内存一定要8G以上。
- 不管是官网部署的还是上面方法部署的OB，就是玩玩。可以跑性能看看，但不要作为评估的结论。毕竟用好OB之前至少得先确保熟悉原理和性能调优方法，可以看看本公众号的一些文章总结。（公众号搜索：OceanBase技术闲谈）。
如果您有任何疑问，可以通过以下方式与我们进行交流：
钉钉群：33254054
github：https://github.com/oceanbase/oceanbase
gitee：https://gitee.com/oceanbase
博客问答：https://open.oceanbase.com/answer
