Re: [I] [Question] nginx负载均衡下插入点边报错 [incubator-hugegraph]

Wed, 23 Jul 2025 23:37:36 -0700 


zjzhaojian commented on issue #2835:
URL: 
https://github.com/apache/incubator-hugegraph/issues/2835#issuecomment-3112228349

   > > 我也遇到了类似的问题, 同样也测试过hstore+pd模式 和 cassandra两种后端, 发现使用cassandra性能更好一点, 
想进一步提升Hugegraph的读写性能, 有哪些Hugegraph和Cassandra配置参数上的调优建议?
   > 
   > [@zjzhaojian](https://github.com/zjzhaojian) 应该不至于吧, 估计是哪里设置不太对, RocksDB 
单节点下随机写入速度就至少几十万/s, 3节点 Store 下可以过百万/s
   > 
   > 如果用高性能的 RocksDB 实现版还能再快 3 倍(已经基本适配 OK 了, 可以内测), 读也是类似的, 差异应该还是挺大的
   > 
   > 你简单的可以先用 twitter 测试数据 + RocksDB 复现一下看看是否符合预期
   > 
   > <img alt="Image" width="791" height="704" 
src="https://private-user-images.githubusercontent.com/17706099/469689814-696d40b6-ffcc-4582-a1e3-df683dcc9fae.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.NjWREAm8oG6ebFhGltODt6EQj57lB6Miu8b5nVsMVfE";>
   
   
   
   > > 我也遇到了类似的问题, 同样也测试过hstore+pd模式 和 cassandra两种后端, 发现使用cassandra性能更好一点, 
想进一步提升Hugegraph的读写性能, 有哪些Hugegraph和Cassandra配置参数上的调优建议?
   > 
   > [@zjzhaojian](https://github.com/zjzhaojian) 应该不至于吧, 估计是哪里设置不太对, RocksDB 
单节点下随机写入速度就至少几十万/s, 3节点 Store 下可以过百万/s
   > 
   > 如果用高性能的 RocksDB 实现版还能再快 3 倍(已经基本适配 OK 了, 可以内测), 读也是类似的, 差异应该还是挺大的
   > 
   > 你简单的可以先用 twitter 测试数据 + RocksDB 复现一下看看是否符合预期
   > 
   > <img alt="Image" width="791" height="704" 
src="https://private-user-images.githubusercontent.com/17706099/469689814-696d40b6-ffcc-4582-a1e3-df683dcc9fae.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3NTMzMzg1OTQsIm5iZiI6MTc1MzMzODI5NCwicGF0aCI6Ii8xNzcwNjA5OS80Njk2ODk4MTQtNjk2ZDQwYjYtZmZjYy00NTgyLWExZTMtZGY2ODNkY2M5ZmFlLnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNTA3MjQlMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjUwNzI0VDA2MjQ1NFomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTEwYWI0ODM1YTUxN2MyOGQ3MWJkOTY5YTkzODYxOWZjZjRlZjFmMzA4OGJjMTRiNGNmYWU4OWIyYWE1NGM1MjgmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0In0.xKbfV-XmBnaCCwdxSAvhU3rvUvmqcdqhu7X0XEo7Hzc";>
   
   在这之前尝试部署的3server+3pd+3store (但是都部署在同一台机器上, 单机多实例),   其他硬件条件: HHD盘, 8C cpu, 
10000M网卡 , 总共120万顶点,  通过rest api提交查询语句 g.V().count()  总共花大约20s,   
对比官方Twitter数据集4千万顶点查询只需2.6s,   是哪些地方导致差距这么大? 


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to