nebula storaged占用内存大小不正常

  • nebula 版本:3.1.0
  • 部署方式:分布式
  • 安装方式:deb
  • 是否为线上版本:N
  • 硬件信息
    • 磁盘 机械
    • CPU、内存信息 Inter 12G

一个三台机器,每台内存都为12G,问题是nebula-storaged的,
本来以为是对应的边或节点太多导致内存变化有折线型,高时占满内存,之后瞬间释放,再一点点长到最高,来回反复.
现在已经把边或顶点都删除没了,再还是会有两台机器,有这样内存变化,这是什么原因引起的?

image

meta / storage / graph info 日志信息
nebula-storaged.INFO

I20220725 11:14:33.583889 30740 NebulaStore.cpp:430] [Space: 1, Part: 4] has existed!
I20220725 11:14:33.583907 30740 NebulaStore.cpp:430] [Space: 1, Part: 5] has existed!
I20220725 11:14:33.583957 30740 NebulaStore.cpp:78] Register handler...
I20220725 11:14:33.583977 30740 StorageServer.cpp:228] Init LogMonitor
I20220725 11:14:33.584151 30740 StorageServer.cpp:96] Starting Storage HTTP Service
I20220725 11:14:33.584548 30740 StorageServer.cpp:100] Http Thread Pool started
I20220725 11:14:33.588665 30847 WebService.cpp:124] Web service started on HTTP[19779]
I20220725 11:14:33.588809 30740 TransactionManager.cpp:24] TransactionManager ctor()
I20220725 11:14:33.589432 30740 RocksEngineConfig.cpp:366] Emplace rocksdb option max_bytes_for_level_base=268435456
I20220725 11:14:33.589475 30740 RocksEngineConfig.cpp:366] Emplace rocksdb option max_write_buffer_number=4
I20220725 11:14:33.589494 30740 RocksEngineConfig.cpp:366] Emplace rocksdb option write_buffer_size=67108864
I20220725 11:14:33.589839 30740 RocksEngineConfig.cpp:366] Emplace rocksdb option block_size=8192
I20220725 11:14:33.960386 30740 RocksEngine.cpp:97] open rocksdb on /usr/local/nebula/data/storage/nebula/0/data
I20220725 11:14:33.960810 30740 AdminTaskManager.cpp:22] max concurrent subtasks: 10
I20220725 11:14:33.978258 30740 AdminTaskManager.cpp:40] exit AdminTaskManager::init()
I20220725 11:14:33.978319 30868 AdminTaskManager.cpp:227] waiting for incoming task
I20220725 11:14:46.636606 30803 MetaClient.cpp:3094] Load leader of "10.199.4.234":9779 in 2 space
I20220725 11:14:46.636663 30803 MetaClient.cpp:3094] Load leader of "10.199.4.235":9779 in 0 space
I20220725 11:14:46.636682 30803 MetaClient.cpp:3094] Load leader of "10.199.4.236":9779 in 2 space
I20220725 11:14:46.636693 30803 MetaClient.cpp:3100] Load leader ok

nebula-graphd.INFO

I20220725 11:14:24.567391 30517 Snowflake.cpp:16] WorkerId init success: 2
I20220725 11:14:24.567610 30648 GraphServer.cpp:59] Starting nebula-graphd on 10.199.4.236:9669
I20220725 11:14:45.168731 30645 MetaClient.cpp:3094] Load leader of "10.199.4.234":9779 in 2 space
I20220725 11:14:45.168785 30645 MetaClient.cpp:3094] Load leader of "10.199.4.235":9779 in 0 space
I20220725 11:14:45.168812 30645 MetaClient.cpp:3094] Load leader of "10.199.4.236":9779 in 2 space
I20220725 11:14:45.168824 30645 MetaClient.cpp:3100] Load leader ok
I20220725 11:16:09.031194 30542 GraphService.cpp:76] Authenticating user root from [::ffff:10.199.4.236]:55414
I20220725 11:16:19.486085 30544 SwitchSpaceExecutor.cpp:37] Graph switched to `recommended_system', space id: 2
E20220725 11:16:21.712992 30545 QueryInstance.cpp:137] SyntaxError: syntax error near `status'
I20220725 11:16:29.813932 30545 SwitchSpaceExecutor.cpp:37] Graph switched to `test', space id: 1
W20220725 11:21:04.564250 30647 MemoryUtils.cpp:76] Memory usage has hit the high watermark of system, available: 8.77384e+08 vs. total: 1.25034e+10 in bytes.
W20220725 11:36:04.610188 30647 MemoryUtils.cpp:76] Memory usage has hit the high watermark of system, available: 1.98242e+08 vs. total: 1.25034e+10 in bytes.
W20220725 11:42:44.569708 30647 MemoryUtils.cpp:76] Memory usage has hit the high watermark of system, available: 1.11951e+09 vs. total: 1.25034e+10 in bytes.

nebula-metad.INFO

I20220725 11:49:10.286832 30604 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.236":9779, role = STORAGE
I20220725 11:49:10.368640 30604 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.234":9779, role = STORAGE
I20220725 11:49:16.854323 30604 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.236":9669, role = GRAPH
I20220725 11:49:16.856009 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.235":9669, role = GRAPH
I20220725 11:49:16.885236 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.234":9669, role = GRAPH
I20220725 11:49:20.302580 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.236":9779, role = STORAGE
I20220725 11:49:20.388725 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.234":9779, role = STORAGE
I20220725 11:49:26.866245 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.236":9669, role = GRAPH
I20220725 11:49:26.884183 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.235":9669, role = GRAPH
I20220725 11:49:26.904654 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.234":9669, role = GRAPH
I20220725 11:49:30.317495 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.236":9779, role = STORAGE
I20220725 11:49:30.404623 30602 HBProcessor.cpp:33] Receive heartbeat from "10.199.4.234":9779, role = STORAGE

https://docs.nebula-graph.com.cn/3.2.0/3.ngql-guide/10.tag-statements/2.drop-tag/
可以看下文档中的说明:

此话题已在最后回复的 30 天后被自动关闭。不再允许新回复。