AWS Presales Solutions Architect at Escala 24x7 Inc.
Real User
Top 5
2024-09-26T16:31:00Z
Sep 26, 2024
You can start small with a basic cluster to learn and practice with it. Selecting the most basic and economical cluster type can save you enough money to move forward with the solution or go with a solution in distribution for deployment.
The solution has very competitive pricing. It can be expensive for the first time when you are building your site. Time and the amount of data also take some time to downsize. It would be cheaper than to have a server, but for Plexigos storage, you have to buy a specific size of compute power. Initially, it was more expensive than BigQuery pay-as-you-go, but it got cheaper later. The more data you have, the relative ratio becomes cheaper. It depends on the use case. In AWS, you must invest and understand the setups, such as what kind of servers you need. Then, you can set up your own, which can be very cheap. Redshift can be engineering-focused to set up, which is not ideal. Azure and Google Cloud, are more efficient for data analysts who are not data engineers. But it can be effective once you get used to it and set up a process. If you are utilizing the on-demand stuff, Redshift is the only vendor offering a dedicated service.
I am not sure about the pricing of the tool since that's not my department. I know that Snowflake is cheaper to set up, manage, and store data. On a scale of one to ten, where one is a low price and ten is a high price, I rate the pricing a seven.
Every solution has a cost and comes in different packages. Considering these factors, AWS Glue is on top. Though Amazon Redshift is good, it depends on what kind of business you're trying to do, what type of analytics you need, and how much data you have. For Amazon Redshift, we pay around INR 60,000 annually. The cost factor also depends on the number of existing customers. In addition to the standard licensing fee paid for AWS, we incur a cloud storage cost of around a quarter million for the amount of data. We also have to bear additional costs for data security and cybersecurity, which are well taken care of by Amazon, hence the premium pricing. There are several other features and services provided by Amazon that justify the premium pricing.
Redshift is costly compared to other solutions. It's pay per use. You can have multiple models. You can go for yearly cost, which is a little discounted than the monthly cost. Depending on how much data you process and store, you can have different pricing. There's no fixed cost. All of these are based on how much data you store monthly and how much data you process.
Data Engineer at a tech services company with 51-200 employees
Real User
2021-07-02T07:10:42Z
Jul 2, 2021
The price of the solution is reasonable. According to the RA3 cluster particularly, it provides 128 GB of storage with only four nodes. If you can manage your computations processes with the help of materialized views and proper queries. I think the IP clusters are very useful and overall fair for the price.
At the moment, pricing is a little bit on the higher side, although it depends on the size of the company. If it's a fairly large enterprise that has a defined budget then they may not have too much of an issue with the pricing.
Cloud & Data - practice leader at Micropole Belgium
Real User
2020-07-19T08:15:38Z
Jul 19, 2020
The cost of Redshift ranges from a few hundred dollars a month to thousands of dollars a month, according to the resources that you're going to use, the number of nodes, and the type of nodes. My customers have implementations that cost about $500 a month for a very small one. I also have a customer with a monthly invoice of about $25,000 USD.
In terms of pricing, if you plan to use the product for a small company and you compare the cost to a product like Oracle Autonomous Data Warehouse, then you need to consider potential company growth and whether you may need to expand the system in the future, require partitions, etc. It could be that in the end, Autonomous Data Warehouse is cheaper than Amazon because if you use that product it can fit your processes and run dynamically. This is not possible on Redshift. If you have two different types of servers on Redshift then at some point you'll reach a limit and will need to move to another type of cluster which is not such an easy operation. On the other hand, with Autonomous Data Warehouse, you can add extra terabytes, but that's it. Nothing else.
What is Amazon Redshift?
Amazon Redshift is a fully administered, petabyte-scale cloud-based data warehouse service. Users are able to begin with a minimal amount of gigabytes of data and can easily scale up to a petabyte or more as needed. This will enable them to utilize their own data to develop new intuitions on how to improve business processes and client relations.
Initially, users start to develop a data warehouse by initiating what is called an Amazon Redshift cluster or a set of...
You can start small with a basic cluster to learn and practice with it. Selecting the most basic and economical cluster type can save you enough money to move forward with the solution or go with a solution in distribution for deployment.
The pricing is reasonable. The cost might increase if we process a huge amount of data.
It is an expensive product.
The product is quite expensive.
The solution has very competitive pricing. It can be expensive for the first time when you are building your site. Time and the amount of data also take some time to downsize. It would be cheaper than to have a server, but for Plexigos storage, you have to buy a specific size of compute power. Initially, it was more expensive than BigQuery pay-as-you-go, but it got cheaper later. The more data you have, the relative ratio becomes cheaper. It depends on the use case. In AWS, you must invest and understand the setups, such as what kind of servers you need. Then, you can set up your own, which can be very cheap. Redshift can be engineering-focused to set up, which is not ideal. Azure and Google Cloud, are more efficient for data analysts who are not data engineers. But it can be effective once you get used to it and set up a process. If you are utilizing the on-demand stuff, Redshift is the only vendor offering a dedicated service.
It's on the expensive side.
I am not sure about the pricing of the tool since that's not my department. I know that Snowflake is cheaper to set up, manage, and store data. On a scale of one to ten, where one is a low price and ten is a high price, I rate the pricing a seven.
Redshift is a bit less costly than Snowflake, but the effort justifies the cost for Snowflake.
The solution is available at a mid-range price as compared to other vendors.
I have heard complaints about the solution’s pricing, and thus I rate it as a five.
In comparison to the price of similar solutions, Redshift's cost is the lowest.
Every solution has a cost and comes in different packages. Considering these factors, AWS Glue is on top. Though Amazon Redshift is good, it depends on what kind of business you're trying to do, what type of analytics you need, and how much data you have. For Amazon Redshift, we pay around INR 60,000 annually. The cost factor also depends on the number of existing customers. In addition to the standard licensing fee paid for AWS, we incur a cloud storage cost of around a quarter million for the amount of data. We also have to bear additional costs for data security and cybersecurity, which are well taken care of by Amazon, hence the premium pricing. There are several other features and services provided by Amazon that justify the premium pricing.
The cost will depend on how you set up your warehouse and what kind of data you store. I would rate the pricing as six out of ten.
The product is cheap considering what it provides; I rate it five out of five for affordability.
The initial setup is very smooth and took us about six months to deploy. The cost is comparable to Snowflake.
Redshift is costly compared to other solutions. It's pay per use. You can have multiple models. You can go for yearly cost, which is a little discounted than the monthly cost. Depending on how much data you process and store, you can have different pricing. There's no fixed cost. All of these are based on how much data you store monthly and how much data you process.
The price of Amazon Redshift is reasonable because it depends on the usage that you use and for DWH for the long term.
I am not aware of the licensing terms.
Licensing costs are reasonable.
The price of the solution is reasonable. According to the RA3 cluster particularly, it provides 128 GB of storage with only four nodes. If you can manage your computations processes with the help of materialized views and proper queries. I think the IP clusters are very useful and overall fair for the price.
The licensing is paid on a monthly basis, and I think the pricing is acceptable. They also have a pay-per-go model which is on a per user basis.
Amazon Redshift costs less than Oracle Cloud or Microsoft Azure.
At the moment, pricing is a little bit on the higher side, although it depends on the size of the company. If it's a fairly large enterprise that has a defined budget then they may not have too much of an issue with the pricing.
The cost of Redshift ranges from a few hundred dollars a month to thousands of dollars a month, according to the resources that you're going to use, the number of nodes, and the type of nodes. My customers have implementations that cost about $500 a month for a very small one. I also have a customer with a monthly invoice of about $25,000 USD.
Of course I would advise others to choose Amazon Redshift, as long as pricing is not a concern for them.
In terms of pricing, if you plan to use the product for a small company and you compare the cost to a product like Oracle Autonomous Data Warehouse, then you need to consider potential company growth and whether you may need to expand the system in the future, require partitions, etc. It could be that in the end, Autonomous Data Warehouse is cheaper than Amazon because if you use that product it can fit your processes and run dynamically. This is not possible on Redshift. If you have two different types of servers on Redshift then at some point you'll reach a limit and will need to move to another type of cluster which is not such an easy operation. On the other hand, with Autonomous Data Warehouse, you can add extra terabytes, but that's it. Nothing else.
It's around $200 US dollars. There are some data transfer costs but it's minimal, around $20.