Alibaba Cloud Redis Helps the Technology Behind the Double Eleven Singles Day

The double eleven singles day is in full swing, and Alibaba Cloud Redis (ApsaraDB for Redis original KVStore) has successfully completed the protection of the double eleven singles day Redis. At present, Alibaba Cloud Redis provides a stand-alone version and a cluster version of Redis.

• The stand-alone version of Redis is highly compatible and supports Lua scripting and geolocation calculations.

• The cluster version features high-capacity, high-performance features that break the single-threaded performance limit of Redis single-threaded.

Alibaba Cloud Redis defaults to dual-system hot backup and provides backup and recovery support. At the same time, Alibaba Cloud Redis source team continues to optimize and upgrade Redis, providing powerful security protection. This article will introduce some business scenarios of the double eleven singles day and then introduce them. The actual business scenario will be more complicated than this one.

Million-level relationship chain storage in the micro-Amoy community

The micro-Amoy community carries the social relationship chain of the billion-level Taobao users. Each user has their own list of concerns. Each merchant has its own fan information. The relationship chain carried by the entire micro-Amoy community.

If the traditional relational database model is used to express the above relationship information, the business design will be complicated and a good performance experience cannot be obtained. The micro-Amoy community uses the Redis cluster to cache the storage community's chain of interest, simplifying the storage of information of interest and ensuring a smooth and seamless experience for the double eleven singles day business. The micro-Amoy community uses Hashes to store information about users' attention. The storage structure is as follows, and the following two query interfaces are provided:

• Whether User A has a relationship of interest with User B;

• User A's active relationship list;

Tmall live commentary product cursor paging

When two users are watching the live broadcast of the wireless terminal, they need to refresh the comments corresponding to the live broadcast. There are three main modes:

• Incremental pulldown. That is, a comment that gets a specified number (increment) from the specified position.

• Pull down to refresh. That is, to get the latest specified number of comments.

• Incremental pull-up. That is, the comment that gets the specified number (increment) down from the specified position.

The wireless live broadcast system uses Redis to optimize the service of the scenario, guarantees the success rate of the live comment interface, and can guarantee tps and millisecond rt requests of 5w+ or above. The live broadcast system will write two data for each live broadcast, which are index and comment data respectively. The data of the index data is SortedSet for sorting the comments, and the comment data is stored by using the hashes, and the index data is obtained through the index when the comments are obtained. A list of comments is obtained by reading the hashes after the required index id. 

After the user refreshes the list, the background information needs to be obtained in the background. The process is as follows:

• Get the current index position

• Get index list

• Get comment data

Sort of orders from the rookie document fulfillment center

Double eleven singles day users will generate a logistics order after generating a trade order. After the order processed by the rookie warehouse allocation system, in order to make the warehouse allocation work more intelligently, the decision system will specify the corresponding order fulfillment plan according to the order information, including when to issue the warehouse, when to leave the warehouse, and when. When the information is collected and collected, and when it is delivered, the document fulfillment center will perform the logistics service according to the performance plan for each stage according to the corresponding time. Due to the limited capacity of the warehouse and the distribution, for the limited capacity, the document that is expected to be the earliest operation is the document with the highest priority, so the order needs to be sorted according to the priority before being actually delivered to the warehouse or matched.

The order fulfillment center uses Redis to sort all logistics orders to determine which order is the highest priority.

Conclusion

Redis is an excellent open source database that provides a more convenient data model for developers. Alibaba Cloud Redis is committed to providing stable, high-performance Redis services. Developers are welcome to use the Ali Cloud Redis service.