-
Notifications
You must be signed in to change notification settings - Fork 93
Home
Switch Languages: English Documents | 中文文档
As a leading third-party risk control company in China, Tongdun Technology handles tens of billions of decision-making requests every day. Therefore, in Tongdun's main data storage infrastructure, Redis is widely used as a cache component. During the peak business period, the cluster actually deploys more than a thousand Redis instances, which is bound to bring great challenges to DBA operation and maintenance management and control. In 2018, we promoted the full containerization of stateless applications in the group, and created a cache cloud product that combines cloud-native technology!
The first version of td-redis-operator can be traced back to 2018. The external open source version is the second version. The development time has continued from July 2018 to the present. At present, the Redis clusters of the two centers in Tongdun are all deployed in On the ultra-large kubernetes cluster.
Current scale:
- Redis instance 5000+
- PB level data
- Involving 1000+ real-time online business applications.
- Name: td-redis-operator
- Language: Golang
- Completely based on cloud native technology to realize resource lifecycle management, fault self-healing, HA, etc.
Description:
- Based on Operator open source products, it is completely operated and maintained on kubernetes.
- Two types of RedisCluster and Active/Standby are supported.
You can use the helm
command to install:
$ helm repo add td-redis-operator https://tongdun.github.io/td-redis-operator/charts/td-redis-operator
$ helm repo update
$ helm install [RELEASE_NAME] td-redis-operator/td-redis-operator
For detailed documentation on installation via helm
, see Installation documentation via HELM .
In addition to installing with helm
, you can also install it via the kubectl
command:
$ kubectl apply -f https://raw.githubusercontent.com/tongdun/td-redis-operator/main/deploy/deploy.yaml
$ kubectl apply -f https://raw.githubusercontent.com/tongdun/td-redis-operator/main/cr/redis_cluster.yaml
$ kubectl apply -f https://raw.githubusercontent.com/tongdun/td-redis-operator/main/cr/redis_standby.yaml
For more YAML files, see Installation documentation via YAML .
Click here to view detailed information about AdminGuide.
Click here to view detailed information about FAQ.
You can also seek help in other ways:
- Issue: https://github.com/tongdun/td-redis-operator/issues
- Wiki: https://github.com/tongdun/td-redis-operator/wiki