重新降低失败恢复期间负载流入不平衡
首先，英语不是我的母语，所以我希望您能理解，即使这很奇怪。
我的系统使用DOT-NET CORE WEB服务器。 我使用redis 写/读取分布 我使用Sentinel建立了故障恢复系统。
来自Dotnet Core Web服务器 使用Redis Master编写，并应用从属的读取方差。
但是，在故障测试期间，写入/读取方差仅针对已更改的主。
如果您关闭写入/阅读分发，请仅与主工作以减轻负载并执行故障测试，将负载很好地放入更改的主机中。
监视正在查看zabbix上的intantaneous_ops_per_sec。
总而言之，在使用Haphroxy进行REDIS的故障转移测试期间，
- 具有写入/读取分布，负载仅集中在主上。
- 如果关闭写入/读取方差，则当然，负载仅集中在主上。
我认为在1的情况下，应该在奴隶雷迪斯上有阅读负载，但是我现在在考虑它，因为它不是。
如果您有这种经验，请分享。
即使我尝试搜索谷歌搜索，我也无法搜索。
程序和相关库
- Dotnet Core 3.1 lib-＆gt; stackexchange.redis.extensions 5.5.0
- haproxy 2.1.7
- redis 5.0.7
First of all, English is not my mother tongue, so I hope you understand even if it is strange.
My system uses a dot-net core web server.
I use redis
Write/read distribution as haphroxy
I built a failure recovery system with sentinel.
From dotnet core web server
Write with redis master and read variance with slave are applied.
However, during the failure test, the write/read variance is directed only to the changed master.
If you turn off the write/read distribution, work with only the master to put the load, and perform a failure test, the load is well put into the changed master.
Monitoring is looking at Instantaneous_ops_per_sec on zabbix.
In summary, during the failover test for redis using haphroxy,
- With write/read distribution, load is concentrated only on the master.
- If you turn off the write/read variance, of course the load is concentrated only on the master.
I think in the case of 1, there should be a reading load on slave redis, but I'm thinking about it now because it's not.
If you have this experience, please share it.
I can't search even if I try googling.
Programs and associated libraries
- dotnet core 3.1
lib -> StackExchange.Redis.Extensions 5.5.0
- dotnet core 3.1
- haproxy 2.1.7
- Redis 5.0.7
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论