Exam AWS Certified Data Analytics - Specialty topic 1 question 33 discussion

Vote A. "currently has a cluster of three dense storage nodes." means it is not single-node cluster, both resizing work, but classic resize take 2 hours–2 days or longer, depending on your data's size. Dense Compute (DC) allow creation of very high performance data warehouses using fast CPUs, large amounts of RAM and solid-state disks (SSDs)

The answer is C. by https://aws.amazon.com/redshift/features/?nc1=h_ls accoranding to the question "...combine all the user data and apply complex calculations that require I/O intensive resources", drop A, B sinice "Dense Compute (DC) the best choice for less than 500GB of data". DS2 (Dense Storage) nodes enable you to create large data warehouses using hard disk drives (HDDs).Most customers who run on DS2 clusters can migrate their workloads to RA3 clusters and get up to 2x performance and more storage for the same cost as DS2. by https://aws.amazon.com/redshift/pricing/ "What to expect" section Once you make your selection, you may wish to use elastic resize to easily adjust the amount of provisioned compute capacity within minutes for steady-state processing.

classic resize does not let you change node type ( note that original config for infra is nodetype = DS2 ) meaning option B is out . If you only want to increase the number of nodes of same type , then you can do that with classic resize ( offcource it will create a new cluster ) so in my opinion C is useless as i can use classic storage itself . Between A and D i vote for A since the use case talks about compute more than storage

definitively C

Bing answer B B. Resize the cluster using classic resize with dense compute nodes. Explanation: Option B is correct because classic resize allows you to change both the node type and number of nodes. In this case, switching to dense compute nodes would provide the I/O intensive resources needed for the complex calculations. The classic resize operation also redistributes the data and reclaims the space, which would be beneficial given the additional 4 TB of user data. Other options are not the best solutions for this scenario. For example, Options A and C involve using elastic resize, which allows you to quickly add or remove nodes, but it doesn’t allow you to change the node type. Option D involves using classic resize with dense storage nodes, but this might not provide the I/O intensive resources needed for the complex calculations.

It's important to note that while classic resize (Option B and D) allows for a change in node types (from DS to DC or vice versa), it involves a longer downtime as it copies data from the old cluster to a new one. Elastic resize is faster than classic resize. it allows you to quickly add or remove nodes to the cluster while keeping the cluster online. This minimizes downtime and can be completed in minutes, as opposed to classic resize, which can take several hours or more, depending on the size of the dataset. Dense compute (DC) nodes are optimized for performance-intensive workloads. They offer faster CPUs, increased I/O performance, and higher storage throughput compared to dense storage (DS) nodes. Given the engineering team's need to apply complex calculations that require I/O intensive resources, switching to dense compute nodes will provide the necessary computational power and I/O performance.

Vote A. The questions specifies I/O as being important which goes right to compute nodes. The reccomendation from AWS is to start with Elastic Resize first. The DC node types have enough to store this amount of data and the number of resized nodes won't exceed the limitations of an elastic resize; primarily the either 2x increase or decrease max of a cluster. If you need to 4x or 8x you node size, then you need to use classic. This was also a (similar) question in Jon Bosno's practice tests in tutorialdojos

You need to store 4TB of user data in Redshift (data that changes little) so the size of these datasets would vary little... They will also execute extremely complex queries... They need to adjust to support these requirements. The first point implies continuing with dense storage nodes, but they also need compute nodes. Given that they have storage nodes and need more computation, I would change the node type, increasing this quantity... Then I would opt for the classic resize. C agreed?

I cant understand this question what to ask me? very abstract question. so I choose C . ChatCPT answerd hahaha.

Elastic as faster and dense storage due to size limit6

I believe option C is the most suitable choice. The reason is that dc2 nodes have a relatively limited SSD capacity of 160GB, while your data size is 4TB. Therefore, you'll need to opt for a dense storage node type to handle the increased storage requirements effectively.

A is correct: Elastic resize across node type automates the steps of taking a snapshot, creating a new cluster, deleting the old cluster, and renaming the new cluster into a simple, quick, and familiar operation. Elastic resize operation can be run at any time or can be scheduled to run at a future time. Customers can quickly upgrade their existing DS2 or DC2 node type-based cluster to the new RA3 node type with elastic resize.

Vote for B. Following is the experience I tried on Sept 1st 2023 - Redshift currently allows me to create a dc2 or ra3 cluster. Although it shows the ds2 option on Resize option, the console will display “NumberOfNodesQuotaExceeded You do not have access to node type ds2.xlarge. Choose another node type” even the quota in the service is enough. In short, I don’t think the candidate can simulate this at home unless he/she currently works for some big companies with existing ds2 Redshift clusters.

The Answer is C: Because... Dense storage nodes provide a balance between "storage capacity and computational power", making them suitable for analytical workloads that involve heavy I/O operations.

Elsatic resize allows to change node types.

Classic resize is for change the node types, we already got storage and we need compute

Classic Resize can allow changing node type and no of nodes, + there is no time constraints added in the question.