Nebula集群新增磁盘后

问答 问题
1

  • nebula 版本：3.2.0

  • 部署方式：分布式

  • 安装方式：RPM

  • 是否为线上版本：Y

  • 硬件信息

  • 问题的具体描述
    我们首次部署Nebula在配置storage的data_path时 只添加了一块磁盘 如：–data_path=/data/nebula/data/storage/ 后面新增了两块磁盘 想提升合并或者写入的性能 配置成了
    –data_path=/data/nebula/data/storage/,/datanebula02/nebula/data/storage/,/datanebula03/nebula/data/storage/ 现在发现 除了最开始配置的路径存入了数据 新增的两个路径 没有存入数据 需要解决方式 如何让3块磁盘均衡使用 另外在使用BALANCE DATA时出现了问题
    image

  • 相关的 meta / storage / graph info 日志信息（尽量使用文本形式方便检索）
    下面是storage的日志
    I20221022 20:18:37.386651 9056 EventListener.h:21] Rocksdb start compaction column family: default because of LevelL0FilesNum, status: OK, compacted 13 files into 0, base level is 0, output level is 1
    I20221022 20:18:37.386703 9056 CompactionFilter.h:92] Do default minor compaction!
    I20221022 20:18:46.375991 9056 EventListener.h:35] Rocksdb compaction completed column family: default because of LevelL0FilesNum, status: OK, compacted 13 files into 7, base level is 0, output level is 1
    I20221022 20:18:46.451633 9056 EventListener.h:158] Stall conditions changed column family: default, current condition: Normal, previous condition: Delayed
    I20221022 20:18:46.451861 9056 EventListener.h:21] Rocksdb start compaction column family: default because of LevelMaxLevelSize, status: OK, compacted 9 files into 0, base level is 1, output level is 2
    I20221022 20:18:46.451903 9056 CompactionFilter.h:92] Do default minor compaction!
    I20221022 20:18:48.834221 9058 EventListener.h:158] Stall conditions changed column family: default, current condition: Delayed, previous condition: Normal
    I20221022 20:19:04.025022 9056 EventListener.h:35] Rocksdb compaction completed column family: default because of LevelMaxLevelSize, status: OK, compacted 9 files into 10, base level is 1, output level is 2
    I20221022 20:19:04.132109 9056 EventListener.h:158] Stall conditions changed column family: default, current condition: Normal, previous condition: Delayed
    I20221022 20:19:04.132360 9056 EventListener.h:21] Rocksdb start compaction column family: default because of LevelMaxLevelSize, status: OK, compacted 6 files into 0, base level is 2, output level is 3
    I20221022 20:19:04.132401 9056 CompactionFilter.h:92] Do default minor compaction!
    I20221022 20:19:04.300324 9058 EventListener.h:158] Stall conditions changed column family: default, current condition: Delayed, previous condition: Normal
    I20221022 20:19:14.090178 9056 EventListener.h:35] Rocksdb compaction completed column family: default because of LevelMaxLevelSize, status: OK, compacted 6 files into 6, base level is 2, output level is 3
    I20221022 20:19:14.148103 9056 EventListener.h:158] Stall conditions changed column family: default, current condition: Normal, previous condition: Delayed
    I20221022 20:19:14.148344 9056 EventListener.h:21] Rocksdb start compaction column family: default because of LevelL0FilesNum, status: OK, compacted 12 files into 0, base level is 0, output level is 1
    I20221022 20:19:14.149070 9056 CompactionFilter.h:92] Do default minor compaction!
    I20221022 20:19:17.790271 9058 EventListener.h:158] Stall conditions changed column family: default, current condition: Delayed, previous condition: Normal
    I20221022 20:19:21.767231 9056 EventListener.h:35] Rocksdb compaction completed column family: default because of LevelL0FilesNum, status: OK, compacted 12 files into 5, base level is 0, output level is 1
    I20221022 20:19:21.831905 9056 EventListener.h:21] Rocksdb start compaction column family: default because of LevelMaxLevelSize, status: OK, compacted 5 files into 0, base level is 3, output level is 4
    I20221022 20:19:21.831952 9056 CompactionFilter.h:92] Do default minor compaction!
    I20221022 20:19:27.595157 9056 EventListener.h:35] Rocksdb compaction completed column family: default because of LevelMaxLevelSize, status: OK, compacted 5 files into 5, base level is 3, output level is 4
    I20221022 20:19:27.635826 9056 EventListener.h:158] Stall conditions changed column family: default, current condition: Normal, previous condition: Delayed
    I20221022 20:19:27.636034 9056 EventListener.h:21] Rocksdb start compaction column family: default because of LevelMaxLevelSize, status: OK, compacted 5 files into 0, base level is 3, output level is 4
    I20221022 20:19:27.636072 9056 CompactionFilter.h:92] Do default minor compaction!

2

只有新建的space才会用上所有盘，已经建的space只会用当时的data_path

1 个赞
3

嗯嗯 自己做了测试 发现是这样的 那么有没有什么方式可以在space之间迁移数据呢 Flink貌似可以的 请问在哪里可以找到示例代码 git上只看见了Flink写入Nebula

4

现在应该不支持space之间迁移数据

5

那Nebula对于生产上新增磁盘是如何应对的呢

6

这个应该只会加在企业版上面。。。

7

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。