修改了OCP的参数ocp.analyze.enabled从false到true,重启OCP后OCP起不来。

【 使用环境 】 测试环境,OCP为企业版
【 OB or 其他组件 】有OAT,OCP,Oceanbase 集群。
【 使用版本 】OCP版本为4.3.5;Oceanbase 为4.2.1 企业版,OCP是单机部署,不是集群。
【问题描述】OCP服务起不来
【复现路径】修改了OCP的参数ocp.analyze.enabled从false到true,重启OCP后OCP起不来。之前修改过ocp参数 ocp.analyze.ob.trace.enabled ,从容器中停止服务,启动ocp服务是能起来的。
【附件及日志】ocp_meta 这个租户所在的容器集群是没有问题的,可以登录的:

摘要

obclient -h127.0.0.1 -P2881 -uroot@ocp_meta -p -A -c
Welcome to the OceanBase. Commands end with ; or \g.
Your OceanBase connection id is 3221590222
Server version: OceanBase 2.2.77 (r119010042024120219-cea8043b056d0429c89d2227eb8b0d965496957b) (Built Dec 2 2024 20:13:21)

Copyright (c) 2000, 2018, OceanBase and/or its affiliates. All rights reserved.

Type ‘help;’ or ‘\h’ for help. Type ‘\c’ to clear the current input statement.

obclient(root@ocp_meta)[(none)]>

进到容器里面,可以看到ic-server.0.out 显示这个服务被disable了。

摘要

[root@OCPserver logs]# ll ic*
-rw-r–r–. 1 admin admin 0 6月 23 22:36 ic-server.0.err
-rw-r–r–. 1 admin admin 360 7月 10 14:50 ic-server.0.out
[root@OCPserver logs]# tail ic-server.0.out
2025-06-23 22:36:56 local ic-server disabled
2025-06-23 22:42:02 local ic-server disabled
2025-07-09 12:19:54 local ic-server disabled
2025-07-09 12:23:40 local ic-server disabled
2025-07-10 12:46:36 local ic-server disabled
2025-07-10 12:54:08 local ic-server disabled
2025-07-10 13:06:18 local ic-server disabled
2025-07-10 14:50:57 local ic-server disabled

ocp.server 日志中的报错信息Unrecognized option: --cp.analyze.opensearch.hosts=127.0.0.1:9200.
Use -h for usage

摘要

[root@OCPserver logs]# ll -lrt
total 558948
-rw-r–r–. 1 admin admin 0 6月 23 22:36 ic-server.0.err
-rw-r–r–. 1 root root 52428941 7月 9 14:57 ocp-server.0.out.10
-rw-r–r–. 1 root root 52428829 7月 9 17:10 ocp-server.0.out.9
-rw-r–r–. 1 root root 52429028 7月 9 19:29 ocp-server.0.out.8
-rw-r–r–. 1 root root 52429180 7月 9 21:37 ocp-server.0.out.7
-rw-r–r–. 1 root root 52429061 7月 9 23:47 ocp-server.0.out.6
-rw-r–r–. 1 root root 52435637 7月 10 02:02 ocp-server.0.out.5
-rw-r–r–. 1 root root 52429620 7月 10 04:17 ocp-server.0.out.4
-rw-r–r–. 1 root root 52429024 7月 10 06:32 ocp-server.0.out.3
drwxr-xr-x. 4 admin admin 4096 7月 10 07:47 obproxy
-rw-r–r–. 1 root root 52429762 7月 10 08:48 ocp-server.0.out.2
drwxr-xr-x. 3 admin admin 4096 7月 10 08:56 ocp
-rw-r–r–. 1 root root 52429631 7月 10 11:01 ocp-server.0.out.1
drwxr-xr-x. 2 admin admin 4096 7月 10 12:46 sandbox
drwxr-xr-x. 2 admin admin 4096 7月 10 14:15 security
-rw-r–r–. 1 admin admin 360 7月 10 14:50 ic-server.0.out
-rw-r–r–. 1 admin admin 410291 7月 10 16:12 ocp_supervisord.log
-rw-r–r–. 1 root root 47302939 7月 10 16:12 ocp-server.0.out
-rw-r–r–. 1 admin admin 237280 7月 10 16:12 ocp-server.0.err
[root@OCPserver logs]# tail ocp-server.0.err
parse option args failed: org.apache.commons.cli.UnrecognizedOptionException: Unrecognized option: --ocp.analyze.opensearch.hosts=127.0.0.1:9200.
Use -h for usage
parse option args failed: org.apache.commons.cli.UnrecognizedOptionException: Unrecognized option: --ocp.analyze.opensearch.hosts=127.0.0.1:9200.
Use -h for usage
parse option args failed: org.apache.commons.cli.UnrecognizedOptionException: Unrecognized option: --ocp.analyze.opensearch.hosts=127.0.0.1:9200.
Use -h for usage
parse option args failed: org.apache.commons.cli.UnrecognizedOptionException: Unrecognized option: --ocp.analyze.opensearch.hosts=127.0.0.1:9200.
Use -h for usage
parse option args failed: org.apache.commons.cli.UnrecognizedOptionException: Unrecognized option: --ocp.analyze.opensearch.hosts=127.0.0.1:9200.
Use -h for usage
[root@OCPserver logs]#

supervirosctl status 显示ic-server_00是EXITED状态,而ocp-server在restart loop:

摘要

[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 02:51 PM
ocp-server:ocp-server_00 STARTING
[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 02:51 PM
ocp-server:ocp-server_00 STARTING
[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 02:51 PM
ocp-server:ocp-server_00 RUNNING pid 97505, uptime 0:00:02
[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 02:51 PM
ocp-server:ocp-server_00 STARTING

日志怎么上传的?已经收集了,没找到怎么上传。

[@论坛小助手]

1 个赞

该问题预期行为,需要搭建数据中台服务,并输入相应地址的,然后ocp.analyze.enabled和ocp.analyze.ob.trace.enabled才能设置为true。

如何更改回原默认值?

登陆ocp_meta租户的meta_database数据库修改表config_properties的相应数据,sql如下

obclient -hxxx -P2881 -uroot@ocp_meta -p'xxx' -Dmeta_database -A
update config_properties set value = false where `key`='ocp.analyze.enabled';
update config_properties set value = false where `key`='ocp.analyze.ob.trace.enabled';  --如果这个参数没修改,这个不动
1 个赞

非常感觉及时帮助。
有两个问题:
一是数据库meta_database不存在,是用ocp代替吗?

摘要

Copyright (c) 2000, 2018, OceanBase and/or its affiliates. All rights reserved.

Type ‘help;’ or ‘\h’ for help. Type ‘\c’ to clear the current input statement.

obclient(root@ocp_meta)[(none)]> show databases;
±-------------------+
| Database |
±-------------------+
| oceanbase |
| information_schema |
| mysql |
| test |
| ocp |
| backup1472 |
| backup147x |
| backup21 |
| backup2230 |
±-------------------+

我已经改完了。执行supervisorctl start ic-server:ic-server_00重启了一下,还是没有效果。

摘要

[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 04:54 PM
ocp-server:ocp-server_00 RUNNING pid 133777, uptime 0:00:02
[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 04:54 PM
ocp-server:ocp-server_00 RUNNING pid 135890, uptime 0:00:03
[root@OCPserver logs]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 04:54 PM
ocp-server:ocp-server_00 RUNNING pid 135957, uptime 0:00:03
[root@OCPserver logs]#

[@论坛小助手]

1 个赞

两个参数都改了吗?

select * from  config_properties  where `key`='ocp.analyze.enabled';
select * from config_properties  where `key`='ocp.analyze.ob.trace.enabled'; 

如果是企业版,建议联系对应支持吧

1 个赞

我的ocp是4.3.5版本,是不是meta_database 不一样?

另外,改回参数后,应该用怎样的步骤重启服务吗?

1 个赞

正常该过后直接重启就可以

1 个赞

use ocp,然后执行的update 语句,都改了:
obclient(root@ocp_meta)[ocp]> select * from config_properties where key=‘ocp.analyze.enabled’\G;
*************************** 1. row ***************************
id: 545
key: ocp.analyze.enabled
default_value: false
value: 0
need_restart: 1
application: ocp
profile: default
label: master
visible_level: PUBLIC
description: 是否开启数据中台模块
create_time: 2025-06-23 22:37:05
update_time: 2025-07-10 16:52:13
description_en: Whether enable analyze module or not
is_az_level: 0
type: BOOL
1 row in set (0.002 sec)

ERROR: No query specified

obclient(root@ocp_meta)[ocp]> select * from config_properties where key=‘ocp.analyze.ob.trace.enabled’\G;
*************************** 1. row ***************************
id: 546
key: ocp.analyze.ob.trace.enabled
default_value: false
value: 0
need_restart: 1
application: ocp
profile: default
label: master
visible_level: PUBLIC
description: 是否开启抓取 trace 数据
create_time: 2025-06-23 22:37:05
update_time: 2025-07-10 16:52:25
description_en: Whether enable analyze ob.trace or not
is_az_level: 0
type: BOOL
1 row in set (0.006 sec)

ERROR: No query specified

obclient(root@ocp_meta)[ocp]>

1 个赞

感觉还是restart loop:
[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 RUNNING pid 150834, uptime 0:00:50
ocp-server:ocp-server_00 RUNNING pid 151320, uptime 0:00:03
[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 RUNNING pid 150834, uptime 0:00:51
ocp-server:ocp-server_00 STARTING
[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 RUNNING pid 150834, uptime 0:00:52
ocp-server:ocp-server_00 STARTING
[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 RUNNING pid 150834, uptime 0:00:53
ocp-server:ocp-server_00 RUNNING pid 151389, uptime 0:00:02
[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 RUNNING pid 150834, uptime 0:00:54
ocp-server:ocp-server_00 STARTING
[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 RUNNING pid 150834, uptime 0:00:55
ocp-server:ocp-server_00 RUNNING pid 151458, uptime 0:00:01

具体要看下ocp server日志

链接:https://caiyun.139.com/w/i/2oxwDE3wcTp0d
提取码:t1yh
复制内容打开移动云盘PC客户端,操作更方便哦
OCP日志的链接在这里:

谢谢!

我们访问不了云盘,帖子里面可以上传附件,你可以压缩后上传下

1 个赞

一上传就报错,是不是因为文件有大小限制?日志包118MB可以吗?

限制50M,压缩下呢

用obdiag 收的,时间跨度为10分钟的日志,这次是29MB。
observer_log_local_20250710175000_20250710180000_616e54_pid_753.tar.gz (29.5 MB)
帮忙看看,谢谢!

现在8080端口可以访问了。
ocp服务看来是好了。
修改完参数之后,需要重启的方法是:
docker stop 602f46672946
docker start 602f46672946
这样重启完之后,虽然还是看到ic-server:ic-server_00 状态时EXITED的。
但是ocp-server:ocp-server_00 可以维持不重启(没有restart loop)
8080端口也可以访问了。

[root@OCPserver conf.d]# supervisorctl status
ic-server:ic-server_00 EXITED Jul 10 06:03 PM
ocp-server:ocp-server_00 RUNNING pid 8, uptime 0:33:48

ic-server:ic-server_00 状态时EXITED的 有什么影响吗?

    ocp-server:ocp-server_00 RUNNING pid 8, uptime 0:33:48

这个是RUNNING的应该就没问题,如果supervisorctl有问题,可以重启或者重装下