Score:0

Inflow of load upon recovery of redis failure

in flag

My system uses a dot net web server

A failure recovery system was built with a redis sentinel system with HAproxy.

Writing as a REDIS Master on a web server and reading distribution with Slave are being applied.

However, during the failure test, there is a phenomenon in which the write/read variance is concentrated only on the changed master.

If you turn off the master/slave write, read distribution, work with only the master to put the load, and perform a failure test, the load is put into the changed master well.

If you have this experience, please share it.

I can't search even if I try googling.

  • my system info. dotnet core 3.1 web server
    nuget -> StackExchange.Redis.Extensions 5.5.0

    haproxy 2.1.7 Redis 5.0.7

VIP haproxy 1 haproxy 2

HA1,HA2 keepalived check 

redis A
redis B
redis C

sentinel X
sentinel Y
sentinel Z
mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.