An IT firm has implemented caching for its three-tier application using multi-node Memcached clusters as a front-end to Amazon DynamoDB.
The ElastiCache Memcached PHP client is used which is configured in auto-discovery mode.
Recently the operations team has performed the addition of nodes in this cluster.
After adding a node to a multi-node Memcached cluster, many cache misses are observed that are overloading the database. What can be configured to avoid this issue?
Click on the arrows to vote for the correct answer
A. B. C. D.Correct Answer: A.
Cache Keys need to be spread across all nodes in an Amazon ElastiCache for better caching performance.
In a normal scenario, when a new node is added to a cluster, many keys need to be moved, resulting in a cache miss.
To avoid this, consistent hashing can be used to have better caching performance whenever new nodes are added to a cluster.
Option B is incorrect as this parameter will add nodes immediately to the cluster.
This will not have an impact on cache misses.
Option C is incorrect as Write-through is one of the caching strategies which will write data in cache whenever data is written in the database.
This will not impact cache misses when a new node is added to the multi-node cluster.
Option D is incorrect as Since auto-discovery is enabled, changing endpoints in the application is not required.
For more information on Amazon ElastiCache, refer to the following URL,
https://docs.aws.amazon.com/AmazonElastiCache/latest/mem-ug/BestPractices.LoadBalancing.htmlWhen a new node is added to a Memcached cluster, the distribution of cached items across the nodes changes. This means that some keys that were previously cached on one node may now be on a different node, and when a client requests data associated with those keys, it will result in a cache miss. This can cause a large number of requests to be forwarded to the backend database, leading to an overload and potentially negatively impacting the performance of the application.
To avoid this issue, the most appropriate solution is to use consistent hashing. Consistent hashing is a technique used to ensure that keys are distributed evenly across the nodes in a cluster. With consistent hashing, when a new node is added to the cluster, only a small percentage of the cached items need to be moved to the new node. This helps to minimize the number of cache misses and reduce the load on the backend database.
Option B, using the --apply-immediately parameter while adding nodes, allows the new nodes to be added immediately, but it does not address the issue of cache misses.
Option C, using write-through caching, involves writing data directly to the backend database and then caching it in the Memcached cluster. While this can help to reduce cache misses, it also introduces additional latency due to the write operations.
Option D, changing the endpoints in the application for new nodes, is not an appropriate solution as it requires manual intervention and can be error-prone.
In summary, consistent hashing is the most appropriate solution to avoid cache misses and overload on the backend database when new nodes are added to a Memcached cluster.