# Conclusion
1. Why do this study: clean cache 時發生後端service無法寫入redis問題。
2. If 硬解(把slave設成可寫),可能發生**數據不一致**。
3. 最可能解法: 把`STAGING`redis的部署模式調成跟`TEST`一樣。
# Part 1 - stackoverflow
[stack**overflow**-Redis sentinel not connect to master: Error: READONLY You can't write against a read only slave?](https://stackoverflow.com/questions/42810115/redis-sentinel-not-connect-to-master-error-readonly-you-cant-write-against-a)

:::spoiler

:::

---
# Part 2 - GPT
Title: redis disable read-only mode on slave node
GPT: [Disable Redis Slave Read-Only](https://chatgpt.com/share/d5b49c51-f488-48a5-be67-e2c48e003009)

aka. make the slave writable

:::danger
not recommand in a production environment because it can lead to data inconsistency issues.
:::
Action - Update the ConfigMap or Secret

apiVersion: v1
kind: ConfigMap
metadata:
  name: redis-config
data:
  redis.conf: |
    ...
    replica-read-only no
    ...

---
# Part 3 - concept & in action(實戰)
[baeldung-redis sentinel hign availability on Kubernetes](https://www.baeldung.com/ops/redis-sentinel-kubernetes-high-availability)

## (實戰) Highly Available Redis Cluster in Action
:::spoiler
:::

### Keyword
`Bitnami` `Explain`
:::spoiler
:::