OCP怎么启动，怎么看是否启动

OceanBase 技术问题
,
#1

image
image1592×381 14.3 KB

之前部署好的，但网页 ip：8080打不开。

但是 obd web 的 8680 是能打开的

怎么检查 ocp是否开启呢？ 为什么打不开管理界面

#3

看下ocp进程 ps -ef|grep ocp
或者直接执行obd cluster start myocp 如果已经启动则会跳过

1 个赞
#4

重启后就报错了，启动不了

image
image1647×604 24.9 KB

[admin@ocp ~]$ ps -ef|grep ocp
admin 3388 1032 0 06:13 pts/0 00:00:00 grep --color=auto ocp

[admin@ocp ~]$ obd display-trace 0cb75742-410b-11f0-b890-5254004d77d3
[2025-06-04 06:13:36.229] [DEBUG] - cmd: []
[2025-06-04 06:13:36.230] [DEBUG] - opts: {}
[2025-06-04 06:13:36.231] [DEBUG] - mkdir /home/admin/.obd/lock/
[2025-06-04 06:13:36.231] [DEBUG] - set lock mode to NO_LOCK(0)
[2025-06-04 06:13:36.231] [DEBUG] - Get deploy list
[2025-06-04 06:13:36.232] [DEBUG] - mkdir /home/admin/.obd/cluster/
[2025-06-04 06:13:36.232] [DEBUG] - mkdir /home/admin/.obd/config_parser/
[2025-06-04 06:13:36.241] [INFO] ±---------------------------------------------------------+
[2025-06-04 06:13:36.241] [INFO] | Cluster List |
[2025-06-04 06:13:36.241] [INFO] ±------±-------------------------------±----------------+
[2025-06-04 06:13:36.241] [INFO] | Name | Configuration Path | Status (Cached) |
[2025-06-04 06:13:36.241] [INFO] ±------±-------------------------------±----------------+
[2025-06-04 06:13:36.241] [INFO] | myocp | /home/admin/.obd/cluster/myocp | stopped |
[2025-06-04 06:13:36.242] [INFO] ±------±-------------------------------±----------------+
[2025-06-04 06:13:36.242] [INFO] Trace ID: 0cb75742-410b-11f0-b890-5254004d77d3
[2025-06-04 06:13:36.242] [INFO] If you want to view detailed obd logs, please run: obd display-trace 0cb75742-410b-11f0-b890-5254004d77d3

1 个赞
#5

metadb启动失败，提供一份详细的obd日志看看

#6

image
image778×327 8.75 KB

重启后，一直卡这里，日志一直狂涮

下面是日志文件：

observer.log (225.9 KB)

#7

[2025-06-04 07:44:13.915567] WDIAG [STORAGE.TRANS] stop (ob_time_wheel.cpp:427) [3093][observer][
T0][Y0-0000000000000000-0-0] [lt=4][errcode=0] ObTimeWheel is not inited

是不是这个问题啊？

1 个赞
#8

[2025-06-04 07:44:13.908458] WDIAG begin (ob_hashtable.h:914) [3093][observer][T0][Y0-00000000000
00000-0-0] [lt=3][errcode=-4006] hashtable not init, backtrace=0x12435d5c 0x116204cc 0x11418847 0
xfae7886 0x2ba5826f5ce9 0x2ba5826f5d37 0x2ba5826de55c 0x52dde1e
[2025-06-04 07:44:13.908474] WDIAG begin (ob_hashtable.h:914) [3093][observer][T0][Y0-00000000000
00000-0-0] [lt=6][errcode=-4006] hashtable not init, backtrace=0x12435d5c 0x116204cc 0x11418847 0
x11418489 0xfae7955 0x2ba5826f5ce9 0x2ba5826f5d37 0x2ba5826de55c 0x52dde1e
[2025-06-04 07:44:13.908485] WDIAG [COMMON] deregister_cache (ob_kv_storecache.cpp:645) [3093][ob
server][T0][Y0-0000000000000000-0-0] [lt=3][errcode=-4006] The ObKVGlobalCache has not been inite
d, (ret=-4006)

error code 4006

1 个赞
#9

这里看到的日志都是早上7.44的。你刚才启动报错的日志麻烦提供一下

1 个赞
#10

我刚同步了时间，安装了ntp，只是时区的设置，时间应该是对的。

启动日志在哪里？

election.log (3.1 MB)
rootservice.log (277.9 KB)

#11

就是observer日志，cat observer.log* |grep "NOTICE"看看

1 个赞
#12

image
image1647×647 42.5 KB

现在全是重复，快速 涮这种日志

#13

notice.log (6.4 KB)

#14

[2025-06-04 08:31:01.796874] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.797200] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.797363] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.797560] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.797756] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.797852] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.798014] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.798339] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.798963] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)
[2025-06-04 08:31:01.798992] INFO [SERVER.OMT] stop (ob_multi_tenant.cpp:593) [7720][observer][T0][Y0-0000000000000000-0-0] [lt=0] there’re some tenants need destroy(count=2)

#15

image
image2474×780 260 KB

提供一下这个时间段的完整observer日志

#16

observer2.log (18.8 MB)
文件太大，上传不了，这个截取了开头的部分

#17

存在5150报错，你的节点之间rpc是否正常。
麻烦提供一下yaml文件在~/.obd/clsuter/xxxx/

1 个赞
#18

inner_config.txt (127 字节)
config.txt (2.0 KB)

OCP部署的单节点的，元数据库的ob也是在这台机器上， 管理的的ob cluster 是另外的。

#19

ob的内存太小了memory_limit: 8GB。或者把ocp的元租户都改为2G。

#20

之前也是这么大，是可以的，重启后就不行了。

现在调到 16G，也还是不行

#21

起不来应该是预期内的，你这个内存配置不合理
sys(最小2G)+meta+monitor=8G了已经。如果无数据的话建议重新搭建一个