1. Course Introduction and Objectives
Through this training, you can systematically master the important knowledge points of Redis and Redis cluster, and quickly improve the actual combat level through expert allocation and experience / case sharing. After each topic, there will be corresponding experiments to practice.
2. Lecturer: Currently working as a database technology consultant for many companies; has done database training for dozens of large enterprises;
joined a foreign company, an Internet, etc. after graduating from the Computer Department of Jilin University;
engaged in consulting on Oracle, MongoDB, MySQL, Hadoop, Redis , Support and training;
providing support and services to a large number of customers in the banking / finance, telecommunications industry, Internet, insurance, etc .;
good at database architecture, problem diagnosis and performance tuning, web diagnosis, big data management, etc.
In 2014, obtained Oracle OCM in
2016, joined the MongoDB core community, engaged in courseware writing and technology sharing in
2018, researched NewSQL, and applied it to online use
III. Class hours: 12 hours
IV. Teaching methods and methods: lectures and case discussions Combining [can flexibly choose
courses that you want to learn and customize] V. Course description:
1: Introduction to ReidsCluster cluster
1. Features of ReidsCluster
2. Server-side sharding
3. Server-side request
4. Request redirection
Two: Redis type overview
- Practice:
Which data type is used in the project
Complexity of each command
Three: Redis development specification
1. Redis development specification design specification and case analysis
Four: RedisCluster deployment and operation and maintenance experience
1. Deployment and precautions
2. Parameter optimization-[affecting service availability]
3. Smooth cluster expansion
4. Overload protection
5. Capacity alarm
6. Fault alarm 7. failover
<tool>
8. Resources Management
(1) Cluster management
(2) Capacity and water level monitoring
9. Shard migration
10. Storage low cost
(1) Hierarchical storage
(2) Separation of cold and hot data
Five: RedisCluster optimization
1. The request volume is too large to fill the machine network card
2. Network card binding
3. slot <access> storm
4. Other optimization parameters
(1) memory related
(2) master-slave synchronization
(3) cluster related
Redis application Scenario analysis 1. The latest list
2. Leaderboard
3. Optimistic lock based on Redis
4. Uniq operation to obtain the re-value of all data in a certain period of time
5. User information management
6. Concerned list
...
Six: Redis distributed service
distribution Advantages and challenges and design priorities
- Service call
Proxy proxy layer design Bao
language support Control
Redis layer connections Security
<authentication, shield dangerous commands>
Batch interface support
Multi-tenant resource isolation
Quota limit - Service governance
Hot key discovery and diagnosis
Big Key optimization ideas
limit limit (key size, automatic elimination)
dynamic compression
expired key
unified setting of key expiration
cache penetration optimization ideas
master-slave DB and cache consistency
distribution Analysis of the Double Write Consistency Scheme of Database and Cache - Monitoring Alarms
Ø application level monitoring
Interface QPS / RT / error rate / hit ratio
Reids monitoring
client connection
cluster status
command calls
memory usage
master-slave synchronization
4. Data Security
Ø redundancy (city, remote)
Ø access security (whitelist mechanism)
Ø operating audit record
seven: RedisCluster practice case studies
eight: Redis data migration
1.redis cluster data migration between [the total amount + increment]
2.redis main migration [the total amount + increment] data from between
3 .redis master-slave migration to redis cluster [full + incremental]
Nine: Redis 4.0 / 5.0 / 6.0 new features
focus on what's new and hot key crawl master-slave replication enhancements and data types, expired algorithms, multi-threaded
ten: city / remote dual live
1. Program Background
2. Objectives Plan
3. city / Brief introduction of the three centers in two places
(1) Application layer active-active and database single-active solutions
Option 1: Multi-node-priority
(2) Application layer active-active, database active-active solutions
Option 1: Dual cluster-data synchronization service
solution Two: dedicated line-add node-priority
scheme 3: active-active scheme for business crossover
- Renew the cache by subscribing to messages
Contact: WeChat 18810049378