【 使用环境 】测试环境
【 OB or 其他组件 】OB
【 使用版本 】OceanBase_CE 4.3.5.0
【问题描述】执行修改log_disk_size参数后查看参数已经生效，但是gv$ob_servers视图中的LOG_DISK_CAPACITY列值并未发生改变，创建新的资源池报错log disk资源不足
【复现路径】alter system set log_disk_size=‘96G’;
【附件及日志】
1 个赞
已解决，再日志中会提示对应的修改资源报错的信息
[2025-07-26 14:15:03.112061] WDIAG [SERVER] decide_disk_size (ob_server_utils.cpp:339) [2311594][OmtNodeBalancer][T0][YB420A00001E-00063ABF9A81977D-0-0] [lt=35][errcode=0] disk_size is greater than total disk space(OB_SERVER_OUTOF_DISK_SPACE=-4184, OB_SERVER_OUTOF_DISK_SPACE=“OB_SERVER_OUTOF_DISK_SPACE”, suggested_disk_size=103079215104, suggested_disk_percentage=0, default_disk_percentage=90, total_space=99703775232, disk_size=103079215104)
[2025-07-26 14:15:03.112102] INFO [SERVER] decide_disk_size (ob_server_utils.cpp:345) [2311594][OmtNodeBalancer][T0][YB420A00001E-00063ABF9A81977D-0-0] [lt=39] decide disk size finished(suggested_disk_size=103079215104, suggested_disk_percentage=0, default_disk_percentage=90, total_space=99703775232, disk_size=103079215104)
[2025-07-26 14:15:03.112143] INFO [SERVER] get_log_disk_info_in_config (ob_server_utils.cpp:88) [2311594][OmtNodeBalancer][T0][YB420A00001E-00063ABF9A81977D-0-0] [lt=39] get_log_disk_info_in_config(suggested_data_disk_size=257698037760, suggested_clog_disk_size=103079215104, suggested_data_disk_percentage=0, suggested_clog_disk_percentage=0, log_disk_size=103079215104, log_disk_percentage=0, total_log_disk_size=99703775232)
[2025-07-26 14:15:03.112210] EDIAG [CLOG] try_resize (ob_server_log_block_mgr.cpp:796) [2311594][OmtNodeBalancer][T0][YB420A00001E-00063ABF9A81977D-0-0] [lt=66][errcode=-4624] try_resize failed, log disk space is not enough(log_disk_size=103079215104, this={dir::"/data/clog/log_pool", dir_fd:16, meta_fd:17, log_pool_meta:{curr_total_size:85899345920, next_total_size:85899345920, status:0}, min_block_id:32145, max_block_id:32862, min_log_disk_size_for_all_tenants_:47244640256, is_inited:true}) BACKTRACE:0x810107f 0x7b89ba6 0x7ee92d4 0x7ee8ac6 0x7ee8768 0x7ee8bfb 0xca6b265 0xca6ad80 0xfc5effd 0x1f9588e5 0x1f9566ee 0x7f192678aac3 0x7f192681c850
[2025-07-26 14:15:03.112240] WDIAG [SERVER.OMT] run1 (ob_tenant_node_balancer.cpp:133) [2311594][OmtNodeBalancer][T0][YB420A00001E-00063ABF9A81977D-0-0] [lt=28][errcode=0] ObServerLogBlockMgr try_resize failed(tmp_ret=-4624)
[2025-07-26 14:15:03.112253] INFO [SERVER.OMT] run1 (ob_tenant_node_balancer.cpp:136) [2311594][OmtNodeBalancer][T0][YB420A00001E-00063ABF9A81977D-0-0] [lt=12] refresh tenant config(tenants=[1, 1003, 1004], ret=0)
[2025-07-26 14:15:03.146358] INFO eloop_run (eloop.c:145) [2311381][pnio1][T0][Y0-0000000000000000-0-0] [lt=59] PNIO [ratelimit] time: 1753510503146347, bytes: 1047617832, bw: 0.013334 MB/s, add_ts: 1007601, add_bytes: 14088
其中default_disk_percentage默认磁盘限制为90%，导致无法使用超过该阈值的空间，修改data_disk_usage_limit_percentage=100;后解决
那个数是不是也设置一下
data_disk_usage_limit_percentage不建议调到最大值，如果遇到磁盘上限问题，将导致集群不可用且无法恢复问题