Amazon Redshift Reviews

It is a great tool since its fully managed. Its console allows you to easily inspect and manage all the queries in addition to the performance of the cluster. It is considered extremely fast querying tool and it allows for concurrent analysis. Amazon RedShift does great with data ingestion and it works on compressing data efficiently. Most of the queries return results fast even with very large data sets. In addition to this, it is considered scalable where if you need more space for your data or in case you want faster results you can add more nodes. The tool has as well a powerful graphical admin monitoring tool that monitors the ongoing queries in real time in addition to historical.

I don't like that the application doesn't support SQL syntax 100% this is sometimes very annoying when you work on developing a query. It doesn't support database keys as well. In addition to this, the workload management mechanism can be done to be more enhanced, dynamic and easy to manage.

Redshift is ideal for your company, especially for small teams. It is a full managed tool that provides you out of the box metrics such as Cloud Watch and it integrates very well with other services provided by AWS. The console for the RedShift is great and the tool itself provides adequate performance for your company. It is easy to learn and use.

We use Amazon Redshift for our data warehouse where our transactional datastores in production are always replicated to Redshift and transformed into fact tables. Redshift main purpose for us is to provide business intelligence and more accurate business processes that the analysts need to be able to make decisions on. It was selected for us for its ability to support big data environment with high availability. It is a great new generation of data warehouse platform for its ability to provide essential business performances data and valuable intelligence. Different departments count on it for a variety of operations and top decision makers rely on it for its great analytics data tools for its ability to track and enhance whenever needed.

Literally takes seconds to setup, integrates with AWS Kinesis via Firehose for huge data volumes and loading. Further you don't need to be a hardcore DBA to manage the system and it works flawlessly with BI Tools like Tableau.

MySQL transition to Postgres for management required some on the job learnings, but nothing insurmountable

Yes I would recommend it to all cloud developers as a good accompanying service provided by amazon

We are in the final stages of moving a 400+ computer co-location facility completely to the AWS cloud switching to over 40 services in AWS stack. During this time we have had 2 account managers and they have both been excellent in providing the AWS resources needed in a timely fashion. Given the significant differences of on premise vs. cloud environments, I am very pleased with the transformation, cost savings and the future products in the pipeline at AWS as our company is continues its AI based solutions for the travel marketplace.

Redshift keeps all of our data together for analytics. We use S3 buckets to collect data and transformdata and then move it to reshift. Looking to to enable with keyword search to make data retrival much easier to use and consume.

Some of the data analytics are difficult based on the side and clarity of data. Looking to enable it to work with keyword search. This will make our data retrevial much easier.

Recommended

Some of the data analytics are difficult based on the side and clarity of data. Benefit is data can be put here on temp basis and upon producing analytics we can remove the data from here. Some of the data analytics are difficult based on the side and clarity of data. Benefit is data can be put here on temp basis and upon producing analytics we can remove the data from here.

I like that redshift allows us to store our large datasets in easily accesible databases. being able to use the postgres syntax i am familiar with makes it easy to adopt. Redshift clusters are quick and easy to deploy.

I find that optimizing databases can be difficult at times. It's a different sort of database to conceptualize based on our previous experiences so understanding the best practices can be challenging. In addition because it is its own database with it's own syntax you can't 100% use postgres syntax or plug-ins directly. It also doesn't have many options for desktop clients.

Storing large datasets for cold storage and legacy uses. This allows us to offset our production databases and keep them lighter, faster and smaller. Being on amazon architecture means that we have reliabilty and automated backups available

I like the large, large amount of raw data you can put in. My company has less than half a terrabyte of nice, simple relational data and redshift obviously meets our size needs.

This is probably due to how we originally designed our star schema, but I find for our most common use cases that the queries we have to write can be much more complicated than queries sent to a non-normalized database. Needing to do window functions is a common need and I think window functions tend to run a bit slower on column storage DBs? Not exactly sure.

We are able to get our valuable data cleaned up and nightly batched into redshift. This enables us to run daily ad hoc queries on all of our clients without querying our production olap database.

Redshift can query huge data sets quickly with little optimization on the developer's side. It integrates well with other systems, like S3 or DynamoDB, and outputs in many different useful formats like Avro or JSON. Since it speaks the PostgreSQL wire protocol, it integrates with basically every tool you would want to use. Unlike most other data warehouses that we evaluated, Redshift is pretty simple to setup and use, and you don't really need to be a BI/DW expert to do so.

Once you hit a certain scale, mistakes you make in the initial design of your tables will become a problem. Since there aren't indexes or anything like that, it might be time consuming to fix mistakes depending on what changes need to be made. There also could be a lot more tools to help you find performance issues. To be fair, lots of performance information is in the documentation, but it's hard to tell that your design isn't as good as it could be until you hit scale.

Fully read the documentation and understand how sortkeys/distkeys work. Also, use the smallest width columns you can as Redshift is more performant at scale whenever the columns are optimally sized. This is different from Postgres, where unbounded VARCHAR columns are faster than fixed length VARCHAR columns.

We needed to be able to quickly analyze data we generate and Redshift was the cheapest and easiest way to do so. Redshift allowed us to get up and running and has given us plenty of headroom to grow and scale.

Amazon Redshift is a robust software. Being a columnar data storage architecture makes it very efficient to query very large amount of data. We have several users who access and query Amazon Redshift but it does not affect the speed or efficency of the queries. It is compatible with several BI and visualization tools. We use Tableau in our organization to create reports and analytics. It is easier to setup and most of maintenance is managed by Amazon. This makes it less resource intensive and easier for your IT team.

Not all features of Postgres SQL are supported by Amazon RedShift. There is no option to create variables which makes it difficult to write complex queries in certain scenarios. It also does not support stored procedures and triggers.

If you have a large amount of data then Amazon RedShift is a good and very efficient alternatrive to other expensive data waerhousing tools.

We have several SQL Server databases which hold data for our different business units. We selected Amazon Redshift as our main analytics data warehouse which store data from all business units (100s of terabyte). We integrate and join data from all sources. We create different data marts which cater to reporting and analytics needs of differnt users in the organization and external clients. We use Tableau as BI tool with our Amazon RedShift data warehouse for most of our reporting & analytics needs.

- Elasticity :

- No specific DBA required - Even application developer can handle . so maintenance cost is low.

- Very very cheap when compare other equilant tool / Databases

- Super fast - Because of the MPP architecture

- There is no key constrains . so these needs to be handled in the ETL or front end.

Easy Environment setup . pay only for your use . Don't think about future usage . whenever you need more space you can add instantly in 2 minutes. Easy maintenance - No separate DBA is required.

Migrating the SQL DW to Redshift . When compare with SQL server configuration, redshift configuration is 50% low, even though we are getting the better performance. Client has saved lot of money through this. They spend almost 2.5 K for one application . Now we are running 4 application for the same cost. .

Redshift's query statistics and tuning tools are critical to the offerring. They enable powerful optimzation opportunities and a create agreat conneciton between the objectives and performance you want to acheive and the neccessary structure and format changes that will drive those wins.

Redshift suffers in the true enterprise DW world due to concurrency issues and slow changes (resizezing or arch changes often have large delays assocaited with them - this is unfortunate. Redshift seems to assume a model of customer's making their data fit the Redshift offering rather than being able to tailor Redshift to the data. As a cloud person I understand this thinking but many potential DW customers in the market see the initial appeal of a managed DW like Redshift only to find that it is not able to meet their performance objects for countless reasons. I am encouraged by improvements like Spectrum

Spectrum is the best thing that happened to Redshift

Data warehousing problems

We've found Redshift to be a great tool for ad hoc, compute-intensive queries. We make extensive use of Redshift's built-in JSON functions, often using these to format data for reporting and analytics purposes.

We've run into a few situations where Redshift's lack of support for a given feature had led to some roundabout solutions (recursive CTEs, cursors, etc.). We also hit Redshift's maximum column width (64k) for some JSON data (even after providing a JSON Paths mapping to the deepest level possible).

We use Redshift as a centralized data store for business intelligence (Tableau, in particular) and analytics.

the powerfull of management all the data also all the configuration of the work management, the easiest tway to configure the database and the easy way to manage the resizing of the cluster.

The extra % of space that AWS provides, that save me a couple fo times, cause we only were able to resize on weekends, and sometimes we reached 100% but we kept writing data in the DB, so thank you aws for that extra space in the cluster

some times the resizing takes a lot of time

The performance it's great, and the easy way to scale up the cluster, you only have to set a couple of setting then your are increasing the cluster automatically

local databases, after migrated to redshift our performance it's much better

The query time and speed is exceptional good even for huge queries of millions of rows

Highly scalable in terms of addition or removal of nodes depending on workload.

Data injestion using S3 Copy command enables insertion of high volume data in seconds.

Availability of sort key and disturbing key helps database design efficiently.

Lack of availability of Update or Upsert facility out of the box.

Limited set of functionality available in terms of data warehouse usage. Support for functions, keys unavailable.

The initiative is to create a centralised data warehouse for the analytical team to generate reports. The analytical tool can be spotfire, Tableau, etc.

Support for native Postgres engine helps the development team as well as analytical team to write queries easily Which reduces the development and testing effort.

Support for AWS S3 copy command to insert data helps the development team to load the data quickly and efficiently.

Redshift is extremely easy to set up. The syntax is similar enough to PostgreSQL that we were able to migrate a lot of our scripts with minimal downtime.

There is enough dissimilarities with postgreSQL that it led to some "gotcha" moments for us. For instance, the default for VACUUM is FULL, whereas on PosgreSQL it's STANDARD (does not recalculate all indexes). This led to some serious I/O hits that we were not expecting.

Do it. It's way better than Vertica and is faster than you think. Columnar storage is the superior architecture. It seriously is lightning fast compared to a standard PostgreSQL database running on RDS with the same specs.

We are attempting to aggregate all of our data into a single repository. Redshift makes it extremely easy to do this since it scales extremely well and is super fast for BI reporting as well as ad hoc querying.

Compared to some data backends used in the past, redshift was considerably faster and scales easily. We were using MySQL before, which has a lot of features, but was taking a lot of time to return data as we got larger.

Some of the SQL features from MySQL aren't there.

It scales and has AWS support. Definitely go through the ducumentation, however, to ensure it has all the SQL functions you may need.

Its a data warehouse that quickly gets me data that I pull when doing a Spark job with EMR.

Redshift enables us to quickly injest data through traditional ETL as well as through Redshift Spectrum, leveraging S3. It gives us MPP horsepower while providing the simplicity of SQL to our less technical staff, enabling them direct query access as well as a variety of BI tools

Can't scale compute and storage independently on the nodes themselves

Quick insights into data. Ability to pull in more data with ease and answer business questions quickly. We've been able to get the performance of query output that our traditional commercial databases could not handle.

Speed. Redshift is blazing fast with seemingly everything. This is its biggest benefit to us.

It has a hard time scaling when you get to big data levels. we have had to migrate some of our datasets into a hadoop/spark cluster to get the performance we needed.

Fantastic Documentation. This has really helped us in our setup and service of this as a product. The fact that everything is easily searchable and readable has really saved us a lot of time when working. Also the admin dashboard is fantastic. Super organized and easy to read and follow.

Redshift primarily is used as a repository for our data that our BI tool reads from. In this sense it is not directly used. However that being said, its speed has really helped us in solving issues quickly.

Fast querying of large datasets. Fast provisioning of hardware. No downtime. SQL interface for our users makes it easy to migrate them. JDBC connectivity allows virtually any client tool to connect.

Certain changes require moving to a new cluster. Write operations are not permitted when you go to encrypt your data in the server.

Also look at Spectrum.

Data warehousing and moving away from legacy DB engine in brick and mortar. Moving to cloud and DevOps allows us to get things accomplished faster. Also, no downtime.

I really liked how easy it was to use and get familiar with it as the code base is a postgreSQL based SQL dialect. The read write cabailities are excellet. Loading data is pretty simple and I liked how most of the error handling is a part of the internals.

A general lack of parameterized SQL statements is a problem in mind. While there are work around for this It seems Jenky to me not to have Stored procedures. I also wish that features like named windows existed and I find it frustrating at time that the things I know about postgreSQL are not always supporeted or are only supported on the leader node.

Right now I am building a datawarehouse with Redshift.

It performas as advertised. Easy to spin up and resize

Scaling clusters is tricky, and more resources and best practice guides would be great. Scaling spectrum or tuning for perfroamcne also seems to be a grey area

Scaling and spectrum are areas that have alot of ambiguity, and support on these fronts would help

We use this as our company data warehouse, and source of truth for analytics. This has unified out companies view on metrics

It is easy to use and reliable. It is currently one of the best solution for handling Analytics. Its a very cost effective solution for small companies which implements analytics.

Sometimes its not able to perform as intended when there is massive data.

We have our analytics supported through the amazon technologies and one of which is the amazon redshift. It is easily scalable most of the times and thats the most important benefit.

I like how simple it is to get started. There is a learning curve in optimization, but it is good that you can get going with very little hassel.

I think it could be made better by keeping up with the features of Postgres a little better. It seems like it is a bit stagnant.

Once you are up and running, review distribution keys, sort keys, indexes and nodes.

Data visibility for analysis by everyone in the company without worrying about scale. We use FiveTran for ETL and Looker for visibility, so our data analysis solution is all SaaS and it is great.

I loved using Redshift at my previous job for analysis. It was speedy and super easy to set up.

We encountered difficulties once we hit our maximum concurrency limit - query queues ended up taking a very long time.

While it works very well, try to undrestand the nuances of using Redshift at scale. It can be done, but I wish I did more research on how to set it up for a user base of 300+ analysts.

We performed Data Analytics on Redshift using SQL. Redshift helped us in the same way our previous analytical database did - provide answers to questions we asked.

Redshift works excellent when querying data of high volumes spanning multi billion data points in each objects. Apart from the standard relational like objects one can create, the addition of Redshift Spectrum enables virtualized views of S3 objects via AWS Glue catalog can be consumed by users.

Redshift database does not work so great for data manipulation activities for updates etc. Also administrative functions like vacuum etc on high volume objects is a time consuming activity.

Used for many purposes in the R&D space, commercial space to store trasnformed schema on write datasets for wider massively parallel consumption.

Its Cloud based and highly scalable, Fault tolerasnt. Provide Python UDF support. Ofers efficeint snapshot solution.

not much, we dont have hugh data sets, so not hitting so many problems.

But one thing I dislike we always have to commit otherwise it wont refresh.

Dont like locks on the table

Created DAtaware house to support Business dashboard. Run metrics calculations.

The query speed for RedShift might be the fastest from all MPP databases available.There are many options to choose from with configurations to optimize your performance.

The scalability IS infinite but NOT instant. When resizing a 40TB database it can take 10 hours+ to resize before the larger or smaller cluster is available.

Redshift is an integral part of my business' data flow and data supply chain. We realize speed and reliability. The Redshift portion is of our data supply chain delivers data to end customers. The reliability of Redshift was the most important aspect.

No hassle , low maintenance

able to deal with PB of data aggregations

Not elastic , to grow clusters u have to

create new clusters and copy over because it is read only when u grow existing

BI analysis was simple and ETL was great to do in RS

SQL queries was fitting for other teams

Ease of use and administration

Performances while taking advantage of the parallelism

Rich SQL language derived from PostgreSQL

Compatibility with existing SQL tools

Monitoring, advanced backup, encryption, column compression

Ease of cluster resizing

Time to business

Costs and maintenance costs

Integration with the Amazon AWS suite

Proprietary solution not installable locally

Creates some dependency with the Amazon AWS suite, for example for the data loading

Not exactly a disadvantage but it is a datawarehouse and not a general propose database

If you put terrible data in, you get terrible data out. It's a lot easier now with redshift to maintain our data integrity.

the copy and unload to file support was great

performance and the space was tightly coupled , some syntax's ware different from psql

we used it for maintaining loges of our product and to store some pre computed values , we never used it for real time computation

Blazing fast parallel processing power.

Top notch compression encodings.

Established data migration workflow, usage with Firehose is a real lifesave.

The `Queries` interface is a bit behind. And their metrics tracking of COPY command is misleading. My experience with COPY is longer than it's reported on the dashboard.

COPY command is top notch. You should build your workflow around this. We made a migration tool pre-Firehose and Firehose just changes the game.

Take a look at compression encodings. It will save up to 80% your disk usage, and faster query time too.

We maintain a duplicate of data from all systems on Redshift to do data warehousing. The performance is just amazing we now cannot live with it. It's managed out of the box and scalable and the computing power and storage algorithms are so good we vouch for this tool everytime anyone talks to us about storing and querying a bulk load of data.

It offers what most other NoSQL databases offer, schema-less data storage. It has built in JSON handling and so much integration with the AWS stack. You can easily setup s3 data stores and easily deploy servers on the AWS platform. Scalability is also a breeze because you can spin up another instance and quickly expand your data offering. Basically it plays nice with the AWS ecosphere and is set up to make the integration seamless. Also it has file integration from s3, I can use it as a document store.

It's only on AWS. That's why I can't really recommend this product unless you know you're going to stay on AWS forever. It really offers no differentiation from any other noSQL store and in fact has less features and reliability over most of the tried and true products in use today. You'd be hard pressed to show where redshift is better than most mature offerings except for the AWS specific integrations.

We used it for storing some statistical analysis of data that we mined from the internet.

It is simple to use and administer. It is easy to deploy cluster.nodes in an cost effective manner.

need more flexibility to create user defined functions. Upgrading/adding new nodes to the existing cluster take a long time and is not a simple task as we need to consider the down-time and resource availability.

It is simple to use and administer. It is easy to deploy cluster.nodes in an cost effective manner. It helps to analyze massive data volumes to get fast and quick response of aggregated data set in seconds.

Analyze massive data volumes to get fast and quick response of aggregated data set in seconds.

Massively parallel computing

Easy to have the cluster up and running in few minutes easy to manage

Easy to scale up and scale down

Easy monitoring

Need optimization to the work load manager queue

Quick commit time will be a great

Business reports and sales targeting

The ability to drill through queries to access data from S3 without needing to replicate the data into the database

a training adjustment for many Oracle developers to understand MPP

RedShift is becoming our ODS in the cloud - providing a secured access layer for users

I like that since moving portions of our data to Redshift, I have not had issues with the server crashing as it happened before.

I mostly just use SQL tables that are being stored in RedShift, so I don't see many negatives

Working towards a better data architecture. The biggest benefit is less downtime with server crashing.

It is very fast if the data is correctly modeled. Easily a replacement for dedicated OLAP engines.

Lag time in refreshing changes to the databases and schemas. 65K limit on varchar columns. Very small compared to other relational databases. Lack of updates to the product feature set. Limited syntax compatibility with PostGreSQL

Beware of the product limitations mainly small 65K varchar column size limit, sparse product updates and limited compatibility with latest PostGreSQL.

We support business dashboards and Tableau reporting with Redshift.

Fast and easy to use. Ability to load massive amounts of data and easily get results.

It's so far behind the current postgres version. I always find myself wanting to use something that isn't supported on redshift.

Data warehousing, analytics, reporting UI

i like the compression algorithm and also the volume it can handle

nothing that i can think of since it serves our use case of reducing query runtime

good performance for specific use cases

queries that are not running in sql server

Amazon Redshift is easy to use, very flexible for data analytics, reporting, data mungling. It connects smoothly with AWS S3, which makes data import/export work seamless.

Sometime tables need to be defined in a certain way to save storage and distribution key needs to be defined explicitly. It is not very user friendly.

Basically we use AWS RS for data processing. we can bring multiple sources of data into one platform and perform data processing at scale.

Fastprocessing, ease of use, postgress is awesome

Compile time is slow on initial queryn - impacts self serve users greatly as their querys are very adhoc

Enterprise data warehouse, BI reporting, selfserve BI, Aggregations

Ease of copying from an S3 data lake saved us lots of development time, and it's ability to scale as our needs grow is huge.

Low number of available databases per cluster

Simple copying from CSV and ORC files already in your S3 data lake is a great benefit.

It’s replaced our SQL Server analytics database for many of our customers. The scalability with respect to performance is really stupid easy.

Fast performance, able to store humungous amount of data, part of AWS, so very easy to integrate with other AWS services such as RDS and S3.

Based on PostgreSQL 8.0.2, which is quite outdated and does not include a lot of new features in newer versions of PostgreSQL.

Understand the use cases and limitations, and decide whether or not RDS is adequate.

Good for large, slow, internal facing queries that analyzes data and provides insights on different aspects of the system.

availablity of database all the time with no maintenance downtime and scalability

performance could be better when the consumption is more

data warehosue on cloud, make it easy scalable and resource independant

The cost vs performance is so much faster than any other solutions I have used.

It is hard to replicate the data off site in real time

Performance without spending a ton

Feels like working on premise, because it's comfortable, easy to understand and even when you dont touch the management in the general overview it seems like you have complete control over the service because of it's good behavior.

Dashboard stills a weird to understand, it should be more intuitive and even you should think about adding graphs that describe to whole db.

Log, product and users data storage

Enhances performance of my platform drastically

Advisors from AWS not easy to comy by. Need an easier way to connect to AWS teams

Performance and scaling

The availability of functions is on-par with older SQL skill sets as well as including important functions such as window functions.

Maybe it is our poor configuration and settings, but Amazon Redshift is an expensive long-term data storage solution that is both slow to query and difficult to structure.

Amazon Redshift is the background database of our BI tool, and for certain uses--like regex() and window functions--we always default to Amazon Redshift.

Managed Service capability with performance and scalability.The feature to Flashback a table from a particular snapshot.

Lack of in-depth documentation on Workload Managment.

Analytics

Columnar storage with fast data retrival

case sensitiveness in join conditions and advanced data processing functions

Business intelligence reporting and data analysis

Performance, ditribution, great storage and easy to spin up

no stored procedures, no auto sacaling.. no constraints

Storing and quering big data

The performance scanning biilions of records

Price model is a little different than market

We used for many realtime data and processing.