VMware Tanzu Data Services Reviews

Guaranteed message delivery, queuing, and low latency delivery.

This allowed us to create a resilient network and independently scale various parts of the system dynamically as the business needs changed.

The biggest area we struggled with was operations troubleshooting. We were running a pretty big cluster and ended up with some random cluster failures that were difficult to troubleshoot. A good portion of these were self inflicted but occasionally the distributed database would end up corrupted.

I have been using RabbitMQ for six years, from prototype to production.

We had a bit of an issue with stability. The usual initial cause would be a hiccup in IOPS in EC2 but then this would cascade into more instability in our main clusters.

For the most part the product was very scalable. The only times we would have problems were usually related to Amazon hiccups that would cause the cluster to slow down.

We did not use their technical support much.

We evaluated a variety of message products and found that for the feature set RabbitMQ was the best.

Initial setup was relatively simple and then we were able to grow into the product by using more of the features.

We used the open source implementation and did not need to pay for support.

We looked at ZeroMQ, Kafka and Redis.

You really need to have or train Erlang expertise. The Erlang tools will become the best way to troubleshoot misbehaving clusters.

We install this solution for our clients. At the moment we are in the middle of an installation for a data warehouse that will be used by a telecommunications company that is based in Lesotho. We have not gone into production yet, but we have used it in a test environment and it works very well.

We are a technology company, so we handle software development, software implementation, data warehousing, and business intelligence.

We are using the on-premise deployment model. In Africa, there isn't much adoption of cloud services, so most of our clients are expecting on-premise implementation.

We chose Greenplum because of the architecture in terms of clustering databases and being able to have, or at least utilize the resources that are sitting on a database.

The installation is difficult and should be made easier. Maybe if the process was simpler it would have a quicker adoption by other developers. This could also be accomplished by providing training aids, such as videos to help with installation or using certain features. There are resources currently available on their website, but you have to search through a lot of documentation.

Our expectation is that the scalability will be good, as it is one of the main reasons that we have invested in this solution.

To this point, I have referenced the material on the website but have not really interacted with technical support.

The initial setup of this solution is not very simple. You need to properly follow the steps in terms of getting the whole architecture put together. We have a team of five people who are working on different aspects of the implementation.

Currently, we are focusing on the data layer. Next will be the ETL layer.

We are using our in-house team to implement this solution for our client.

We have used Oracle and Microsoft SQL, but we haven't had much success. We found that Oracle was not as scalable and we were having some performance bottlenecks. Also, from a licensing perspective, Greenplum was a better choice. For all of these reasons, we have chosen to invest heavily in Greenplum.

I would recommend this solution specifically for the scalability. This solution has a more futuristic technology, as opposed to the old school kind of data warehousing. If people are interested in getting something that is more future-proof, then I would recommend this solution.

So far, we're comfortable with what we've seen. What we have configured is addressing our needs at the moment.

I would rate this solution an eight out of ten.

Legacy queuing systems have been replaced by RabbitMQ. The performance has been increased to a great extent.

The message routing is the most valuable feature. It is effective and flexible.

The debugging capabilities and testing flexibilities need to be improved.

The stability was fine.

There were no scalability issues as such. The scalability was fine.

I would give technical support a rating of 5/10.

Initially, we were using different queuing technologies. Due to the message routing feature and flexibility that RabbitMQ provided, we made the switch to this tool.

The setup was easy enough, once we had done proper research before the implementation.

The pricing is okay.

This product needs to be understood completely before implementing it. One should not be mistaken that it will replace the whole messaging system as such.

Greenplum is an MPP architecture database. Data can be distributed across multiple nodes and strong distribution will allow queries to execute on all segments at once, which is very powerful. As long as we have good SQL knowledge, we can start playing in the platform. Greenplum uses Postgres and ANSI Standard SQL. Also, it supports many other procedural languages, such as Python, C++, and Pearl.

Greenplum is a high powered, multi-node database that was chosen for its capacity to ingest and query data at extremely high rates of speed, enabling in Database Analytics and Statistical output on granular levels of data that was otherwise inaccessible before its deployment. We were able to analyze and produce output on large volumes of data very quickly which saved us lots of time (we used to wait for hours to get the same output). The management was able to get insights very quickly so that they can make informed decisions.

I used Greenplum between Aug 2011 – Aug 2015. Almost all the members in the analytics team used Greenplum on a daily basis.

There were no issues, and it was doing what it was supposed to do.

There were no issues, and it was doing what it was supposed to do.

We had a pre-sales consultant who provided end-end solution about the product. Also, he was working with our data and clearly demonstrated the advantages of Greenplum. After we purchased the product, we were provided a full time consultant who had extensive knowledge about the product. He was primarily responsible for providing hands on experience on projects and also did excellent job of teaching everyone and bringing everyone up to the speed on the new platform. We also had a technical person offshore who was responsible for fixing things if something breaks up or any other issues.

We did use other products in the company but it wasn’t an MPP architecture database. Our data was getting bigger so we needed something with MPP architecture to tackle big data challenges so Greenplum was considered. It was a management decision to purchase this product (not sure whether other similar products were considered or not)

All the set up I believe was done by Greenplum team.

I didn't have any visibility on the pricing and licensing. But I can say that, we needed product like Greenplum to store, manage and analyze huge volumes of data which can be daunting task

Greenplum is a MPP (massively parallel processing) database which is extremely fast. If people are dealing with very high volume of data, it is definitely a product to consider seriously

We use the solution on our SaaS platform to speedup and simplify customer access across services.

Some of the most valuable features are “publish and subscribe”, fanout queueing, and scalability.

We have a number of different use cases in our scenario. A key one is “publish and subscribe”. We have spent the last year breaking up a large monolithic application into microservices and each microservice has to subscribe to different events for the purpose of CQRS and other kinds of updates. RabbitMQ is perfect for “publish and subscribe”. It does an awesome job at fanout, perfect for CQRS, messages are delivered to all subscribers with almost no additional latency.

RabbitMQ provides the ability to scale queues in a very simple and elegant way. If it had a “failure queue” with robust delivery and recovery built-in with the same power, that would be great. We use a completely different queuing system for failures. So there is a little more effort to take messages in a failure queue, analyze them, figure out what went wrong and then restart them in Rabbit. It is doable, and we do it, but if we had a round trip solution in Rabbit, that would be awesome.

For me, having a robust failure queue, is high on the list of improvements needed in the near future. This is an important update needed because right now we are using Doctrine for our failure queue. Doctrine does a great job.

I have been using the solution in the past year.

Rabbit is a very scalable solution. We could easily queue 50,000 messages in less than a minute. The first day we introduced Rabbit to replace another queueing system that we were using, there was disbelief on the part of the product team because the response was so fast. We need tens of thousands of messages queued in a short period of time, approximately one minute. For example, one user action could spawn 65,000 messages. We also need the ability to segregate different queues. This solution did a great job.

The installation is very simple and elegant, and we love the graphics. It lets us see exactly what is happening with the ability to start the queue, stop the queue, consume messages on the queue. This is a huge help.

We design, develop and deploy the solution ourselves.

We are also evaluating Apache Kafka. Our process is very disciplined. We look at the analytics, the abstraction, the architecture relative to our technical architecture, we ask ourselves questions about the use case, which is better for use A or B. Kafka is not as simple for “publish and subscribe”. You can do it, but not the best fit for us. However as a queueing system, Kafka is great. The records are stored on the queue in the order they are received, However, you can easily search by topic no matter how large the list. Important if you keep track of everything.

There are many different use cases for each technology, as well as many approaches. So have the architecture team graph and document every solution. Have a few training days to clarify the goal, the solution and the implementation. One of the things we do in our training is to actually create prototypes, the abstract model of our ideal state. This demonstrates exactly what we all need to do. Developers understand more quickly with a model. It flattens their learning curve and they are more productive more quickly.

I rate VMware RabbitMQ a ten out of ten.

Asynchronous messaging; supporting data integrations between multiple applications on behalf of our many customers. RabbitMQ allows us to elegantly fan-out data to a variable number of subscribers, with almost zero effort.

We have been able to set up a messaging system that facilitates data integration between the software modules that we sell.

RabbitMQ allowed us to do this quickly so that we could focus on the business requirements, rather than divert our efforts to message broker implementations.

Once the architecture was proven, we were able to return to the RabbitMQ message layer in order to implement an HA cluster with a minimum of problems encountered.

Our business now has a fit-for-purpose information hub that we can apply across our systems. As the customer-base grows, we know that the hub can grow with it.

RabbitMQ is a solid, widely-used messaging system with a low cost-of-ownership. It is open, but with commercial support potentially available from Pivotal if required. (We have never needed it.) There is also a strong online user community.

One crucial feature was guaranteed messaging. We needed a solution that we could trust to not lose data.

Its built-in clustering capability allowed us to configure it as a highly available message broker, so that we can have confidence in the resilience of our architecture.

It can be scaled as well, although we have not tested this.

After almost two years' usage in our production environment, I am impressed by how stable the platform is - even when running on Windows Server 2012. Sure, we have had to tweak our set-up here and there as we have learned a few operational lessons along the way but overall it is very good.

RabbitMQ is clearly better supported on Linux than it is on Windows. There are idiosyncrasies in the Windows version that are not there on Linux.

The documentation for the Windows version is also less plentiful and less accurate.

The online community clearly provides better Linux support, but this naturally follows from the smaller Windows installed base.

There are also some potential concerns about how we maintain high-availability whilst also scaling out.

We have had no stability issues.

We have not used the scalability features yet.

We have not used technical support.

No previous solution was used.

The initial setup was straightforward. The online documentation was adequate and there is minimal initial configuration required to get up and running.

After that, it is simply a matter of experimentation with the various features and learning as you go.

This is an open source solution.

We looked at MSMQ, NServiceBus, Azure Service Bus, and Apache Kafka.

I would recommend that anyone who intends to deploy RabbitMQ on Windows should first consider whether a Linux implementation is a viable option for their situation.

The most valuable feature is it’s robustness. Message queues need to be extremely reliable as they are the glue between system components.

Also, the speed is important and its good scaling capabilities.

It allows developers to focus on application functionality without having to re-invent interprocess communication, which is difficult.

I also allows us to develop smaller, more efficient, and less complex subcomponents of a larger application.

I would like to see better documentation on how to set up complex webs of RabbitMQ servers — master/slave, multi-master, etc.

I have been using RabbitMQ for 7+ years.

We have not encountered any stability issues.

We have not encountered any scalability issues.

We were using IBM MQ, but it was too costly and not open source.

The initial setup was simple for my applications, but I have not used RabbitMQ on a complex project that would require clusters of servers.

My advice is to read the message boards and play with the API.