5M1.空间：
[admin@localhost ~]$ df -h |egrep home|data
/dev/mapper/centos-sdd 49G 52M 46G 1% /data/log1
/dev/mapper/centos-sdb 50G 526M 47G 2% /home
/dev/mapper/centos-sdc 50G 52M 47G 1% /data/1
2.内存：
[admin@localhost ~]$ free -g
total used free shared buff/cache available
Mem: 9 0 7 0 1 5
Swap: 6 0 6
[admin@localhost ~]$
3.启动oceanbase进程命令：
bin/observer -i eth0 -P 2882 -p 2881 -z zone1 -d /home/admin/oceanbase/store/obdemo -r ‘192.168.12.161:2882:2881;192.168.12.162:2882:2881;192.168.12.163:2882:2881’ -c 20210122 -n obdemo -o "memory_limit=8G,cache_wash_threshold=1G,row_purge_thread_count=1,net_thread_count=1,enable_syslog_recycle=True,enable_merge_by_turn=FALSE,cpu_count=4,location_refresh_thread_count=2,max_syslog_file_count=3,major_freeze_duty_time=Disable,_easy_memory_limit=1G,schema_history_expire_time=1d,merge_thread_count=1,writing_throttling_trigger_percentage=70,max_trx_size=2M,workers_per_cpu_quota=2,enable_separate_sys_clog=0,datafile_disk_percentage=90,min_full_resource_pool_memory=536870912,system_memory=5G,memory_chunk_cache_size=128M,trace_log_slow_query_watermark=10s,writing_throttling_maximum_duration=1m,switchover_process_thread_count=1,datafile_size=40G,minor_freeze_times=500,stack_size=1536K,disk_io_thread_coun=1,config_additional_dir=/data/1/obdemo/etc3;/data/log1/obdemo/etc2"
4.observer.log error信息
[admin@localhost log]$ pwd
/home/admin/oceanbase/log
[admin@localhost log]$
[admin@localhost log]$ ll
total 876
-rw-r–r-- 1 admin admin 0 Jan 25 15:15 election.log
-rw-r–r-- 1 admin admin 0 Jan 25 15:15 election.log.wf
-rw-r–r-- 1 admin admin 838861 Jan 25 15:17 observer.log
-rw-r–r-- 1 admin admin 39200 Jan 25 15:17 observer.log.wf
-rw-r–r-- 1 admin admin 10461 Jan 25 15:17 rootservice.log
-rw-r–r-- 1 admin admin 0 Jan 25 15:15 rootservice.log.wf
[admin@localhost log]$
[admin@localhost log]$ cat observer.log | grep error
[2021-01-25 15:15:26.587362] INFO ob_server_config.cpp:213 [23534][0][Y0-0000000000000000] [lt=9] | ignore_replay_checksum_error = False
[2021-01-25 15:15:26.587438] INFO ob_server_config.cpp:213 [23534][0][Y0-0000000000000000] [lt=9] | ignore_replica_checksum_error = False
[2021-01-25 15:15:26.588740] INFO ob_server_config.cpp:213 [23534][0][Y0-0000000000000000] [lt=9] | enable_rich_error_msg = False
[2021-01-25 15:15:26.639543] INFO [COMMON] ob_io_manager.cpp:641 [23534][0][Y0-0000000000000000] [lt=19] Success to config io manager, (conf={sys_io_low_percent:0, sys_io_high_percent:90, user_iort_up_percent:100, cpu_high_water_level:320, write_failure_detect_interval:60000000, read_failure_black_list_interval:300000000, retry_warn_limit:2, retry_error_limit:5, disk_io_thread_count:8, callback_thread_count:8})
[2021-01-25 15:17:01.872905] INFO ob_server_config.cpp:213 [23660][0][Y0-0000000000000000] [lt=8] | ignore_replay_checksum_error = False
[2021-01-25 15:17:01.872973] INFO ob_server_config.cpp:213 [23660][0][Y0-0000000000000000] [lt=8] | ignore_replica_checksum_error = False
[2021-01-25 15:17:01.874096] INFO ob_server_config.cpp:213 [23660][0][Y0-0000000000000000] [lt=8] | enable_rich_error_msg = False
[2021-01-25 15:17:01.920714] INFO [COMMON] ob_io_manager.cpp:641 [23660][0][Y0-0000000000000000] [lt=22] Success to config io manager, (conf={sys_io_low_percent:0, sys_io_high_percent:90, user_iort_up_percent:100, cpu_high_water_level:320, write_failure_detect_interval:60000000, read_failure_black_list_interval:300000000, retry_warn_limit:2, retry_error_limit:5, disk_io_thread_count:8, callback_thread_count:8})
[admin@localhost log]$
5.observer.log最后10行
[admin@localhost log]$ tail -10f observer.log
[2021-01-25 15:17:02.953661] INFO ob_timer.cpp:233 [23660][0][Y0-0000000000000000] [lt=15] cancel task(this=0x7fc90c019310, thread_id=23698, pos=0, wakeup_time=6921383093, tasks_num=0, task=tasktype:N9oceanbase6common15ObKVGlobalCache15KVStoreWashTaskE)
[2021-01-25 15:17:02.953684] INFO ob_timer.cpp:233 [23660][0][Y0-0000000000000000] [lt=15] cancel task(this=0x7fc90c019750, thread_id=23699, pos=0, wakeup_time=6921531032, tasks_num=0, task=tasktype:N9oceanbase6common15ObKVGlobalCache16KVMapReplaceTaskE)
[2021-01-25 15:17:02.953703] INFO ob_timer.cpp:105 [23660][0][Y0-0000000000000000] [lt=14] ObTimer destroy(this=0x7fc90c019310, thread_id=23698)
[2021-01-25 15:17:02.953861] INFO ob_timer.cpp:355 [23698][76][Y0-0000000000000000] [lt=57] timer thread exit(this=0x7fc90c019310, thread_id=23698)
[2021-01-25 15:17:02.954056] INFO ob_page_manager.cpp:41 [23698][0][Y0-0000000000000000] [lt=28] unregister pm finish(&pm=0x7fc8d8af6880, pm.get_tid()=23698)
[2021-01-25 15:17:02.956354] INFO ob_timer.cpp:105 [23660][0][Y0-0000000000000000] [lt=14] ObTimer destroy(this=0x7fc90c019750, thread_id=23699)
[2021-01-25 15:17:02.956521] INFO ob_timer.cpp:355 [23699][78][Y0-0000000000000000] [lt=25] timer thread exit(this=0x7fc90c019750, thread_id=23699)
[2021-01-25 15:17:02.956754] INFO ob_page_manager.cpp:41 [23699][0][Y0-0000000000000000] [lt=28] unregister pm finish(&pm=0x7fc8d88d4880, pm.get_tid()=23699)
[2021-01-25 15:17:03.045124] INFO [COMMON] ob_kv_storecache.cpp:232 [23660][0][Y0-0000000000000000] [lt=27] The ObKVGlobalCache has been destroyed!
[2021-01-25 15:17:03.045259] INFO [SERVER] main.cpp:439 [23660][0][Y0-0000000000000000] [lt=51] observer is exit
observer.log里的日志内容很多，不是很好懂。不过真有问题的时候，也能看出一些信息。搜索 ERROR ，找那些“说人话”的日志。
比如说这个日志里有:
“[2021-01-25 15:15:26.639656] ERROR [SERVER] init_io (ob_server.cpp:978) [23534][0][Y0-0000000000000000] [lt=17] init io benchmark fail, (ret=-4027) BACKTRACE:0x7a794ba 0x7a24476 0x410a61 0xf0873e 0x4bc90e4 0x4bd2def 0x44992c6 0x7fd9b5be73d5 0x44f8385
”
这个是常见 ERROR，因为不是 OCP 安装的，没有提前测试磁盘 IO生成相应的 etc\io_resource.conf 文件 ,所以会有这个 ERROR。不过这个不影响 observer 的使用。
“[2021-01-25 15:15:32.187374] ERROR [STORAGE] scan_store_dir (ob_raid_file_system.cpp:672) [23534][0][Y0-0000000000000000] [lt=23] disk num in store dir is not valid(ret=-4337, disk_num=-1, OB_MAX_DISK_NUMBER=32) BACKTRACE:0x7a794ba 0x7a24476 0x412fda 0x30d2695 0x68ac73b 0x68ac893 0x616c3a4 0x616c6c0 0x632080e 0x4bcb495 0x4bd319c 0x44992c6 0x7fd9b5be73d5 0x44f8385
”
这个说明数据目录有问题。
“[2021-01-25 15:15:32.187698] ERROR [STORAGE] init (ob_store_file.cpp:442) [23534][0][Y0-0000000000000000] [lt=53] Fail to init store file system, (ret=-4337) BACKTRACE:0x7a794ba 0x7a24476 0x410a61 0x25f83a0 0x616c9e2 0x632080e 0x4bcb495 0x4bd319c 0x44992c6 0x7fd9b5be73d5 0x44f8385”
这个进一步说明数据目录初始化失败。
那么后面看看数据目录的权限、组织结构以及启动参数是否正确。
数据目录虽然计划是 /data/1 。这个是靠 /home/admin/oceanabase/store/集群名/下各个目录映射到 /data/1/集群名/ 下。
评论
有用 0看日志是在observer init fail，主要是因为初始化存储init_storage失败，从日志里看启动命令
bin/observer -i eth0 -P 2882 -p 2881 -z zone1 -d /ob -r …
这里的-d指定的是/ob好像不是你的数据目录位置
但是从你实际的命令：
bin/observer -i eth0 -P 2882 -p 2881 -z zone1 -d /home/admin/oceanbase/store/obdemo -r ‘192.168.12.161:2882:2881;192.168.12.162:2882:2881;192.168.12.163:2882:2881’ -c 20210122 -n obdemo -o …
看-d /home/admin/oceanbase/store/obdemo
按楼上梅老师的方法，先检查下目录设置和权限等。
另外发现你这个命令 -r 参数后面的单引号是中文状态的，这个不对，要是英文状态的单引号才行。不知道会不会是这个特殊字符导致的问题，你仔细检查下。
评论 有用 0墨值悬赏