Microsoft Parallel Data Warehouse Reviews

We provide this solution as a service to customers. Sometimes, it is just used as a data warehouse. However, a couple of our customers use the data warehouse to collect information from everywhere, e.g., from their subsidiaries, such as electric or petrol companies.

We are also developing some dashboards using the data warehouse. Sometimes, you are using Microsoft BI, but most of the time, you are using Qlik. It depends on the customer needs at the end of the day. 

Most of our limitations until this moment have been on-premise. There is some sensitive information, like the billing system for an entire country's electricity. However, because of the current circumstances with COVID-19, we are discussing with them how to move it to the cloud. Mostly because of the regulation that any government data cannot be hosted on a cloud, we need to host it on a local cloud at the moment. We cannot use AWS or similar things as of now. 

While we do have local cloud vendors that we are working with, this is something in progress.

• Availability of resources everywhere
• The cost of resources
• The information knowledge is there.
• It is easy to use. You can do what you want with the tool.
• It integrates well with SQL Server.

Collecting the data through SSIS packages from different sources and putting them all in one data repository is the most powerful thing. While others have this feature, they don't have the simplicity or ease of use when getting a resource and knowing everything about it. You can collect data from Oracle Databases, SQL databases, or file systems where it is then saved in one location, building cubes in an easy for SSIS packages. 

The setup has room for improvement. Some customers want to go to non-Windows Servers. Have the entire SQL Server done there though is an obstacle for us, as the configuration of Microsoft BI with cubes is a bit complicated.

I would like the integration to be offered in a simpler way. I would like improvement in the integration between Microsoft SQL cubes and Parallel Data Warehouse with other members of the Microsoft family, such as Microsoft Power BI and SharePoint.

This tool needs a lot of memory for processing. As it does a lot of jobs, it may need this memory while other tools don't need that much. It would help if they enhanced the memory and hardware usage of the tool in order to support the performance of doing the queries

I would like the tool to support different operating systems.

I want a dashboard.

We have been using it for five to six years.

It is stable. We have migrated millions to billions of data points in Cubes from system to system, while running data cleansing conditions. Putting restraints on this can be time consuming which is why people may want to go to another server. When you inject equation and conditions, then it consumes a lot of time. Though, I have never compared this time consumption to other tools.

I have no concern about the scalability. We have never faced any problems regarding the scalability. If we wanted to extend this, it is extendable.

The online help is good enough along with our technical expertise. 

If everything is available, such as information knowledge sources and connections, I can have the solution deployed in five days.

I recommend getting assistance from an integrator or consultant with the integration.

Microsoft has an agreement with the government in our country, so our customers get their licensing costs from the Ministry. Whenever we work with any government, company, or government institute, which is mainly what we are doing, that license comes directly from the Ministry of Technology and Information. 

We actually do not know that much about the price of Microsoft tools. Even if I ask for the price, I'm asking it as a partner.

We have had a good experience working with this tool.

If you don't have a problem with memory and have good processors, then you won't have an issue. However, depending on your needs, you may want to go check out other tools.

I would rate the tool as an eight out of 10.

On-premises

I am using the Microsoft Parallel Data Warehouse for working with information services and datasets, creating reports. It is used primarily for our historical data, but we do some in real-time also. We can receive data weekly, nightly, or even hourly.

The most valuable feature of this solution is performance. It reduces the time it takes to process data.

The initial setup of this solution is complex and can be improved.

Sometimes you need to re-run the reports because they don't work.

The reporting for certain types of data needs to be improved.

I have been working with these kinds of services for the past five to ten years.

The stability depends on your data and whether it is null. If the data is set up properly then it will be more stable.

It is definitely scalable. We have more than 220 users in multiple cities. Most of the people, perhaps 70%, are information technology users. The remaining 30% are business analysts who are concerned only with reports.

We do plan on increasing our usage. We recently converted all of our Tableau data, and now we are working on permissions and security.

The technical support from Microsoft is good.

The initial setup is definitely complex.

We are not always aware of where the data comes from because it is distributed. Sometimes the data is fine and sometimes it needs to be set up. Once it is configured then it runs fine in parallel.

The length of time for deployment depends on the size of the data and whether we were also implementing a data warehouse or a data mart. When I was implementing this solution and copying data, it took between six and eight months. That is end-to-end with everything finished.

If it is a small data mart then it may only take a month.

Our in-house team, including our DBA, handled the deployment.

My advice for anybody who is implementing Microsoft Parallel Data Warehouse is to be aware of how big the data is. You have to be sure of how to secure and take care of the data. You have to think about scalability.

I would rate this solution a seven out of ten.

This is a solution that we implement for our clients. We provide end-to-end development, maintenance, enhancement, testing, and support. Everything is done by us.

We are able to monitor daily jobs, so if there is anything that needs to be done then we can do it. If there are discussions with our clients about changes that need to be made then it is us who implements them.

The most valuable feature is that we can deploy directly to the production environment, rather than having to go through a development environment first.

If the database is large with a lot of columns then it is difficult to clean the data.

I have been using the Microsoft Parallel Data Warehouse for two years.

The Microsoft Parallel Data Warehouse is a stable solution.

We have not had problems with scalability.

The initial setup is not difficult. It is easy and we are fine with that.

We deploy this solution for our clients.

I am not comfortable with the Microsoft Parallel Data Warehouse at the moment.

I would rate this solution a six out of ten.

We are using PDW as an EDW solution.

It helped, initially, as a replacement for our DW DB, but later on faced issues due to concurrency, which forced the customer to use the DB as a secondary solution.

Nothing specific, comparable to other solutions.

Concurrent queries are limited to 32, making it more of a data storage mechanism instead of an active DWH solution.

They need to improve the metadata being captured to a greater duration.

One to three years.

MPP processing gives us the ability to distribute large data sets across nodes.

We delivered a data warehouse for Contactless and Oyster at TFL.

Improve the speed of processing replicated tables.

We have been using this solution for years.

There were stability issues when the product was in beta.

There were scalability issues in that there is a limit to 32 concurrent queries.

Technical support is good.

We switched from the standard SQL server 2014. We use PDW to improve overall ETL and report performance.

The initial setup was complex. It was fairly challenging to migrate from SQL server to PDW.

Make sure your data volumes are very large: At least 60 million rows per table.

Microsoft SQL Server Parallel Data Warehouse provides significant performance boost from 10-100 X times faster for data loading operation.It minimized the cost of design with lower infrastructure requirements.It is designed to provide optimized performance on latest industry hardware.

It requires significant infrastructure expertise to implement. Possibility of over specified storage or under specified CPU (need proper planning before starting implementation)It requires significant SQL Server expertise. Microsoft SQL Server Parallel Data Warehouse is purposely built to provide high performance data warehousing. By adopting industry standard hardware it avoids vendor lock-in issue. It is best suited for mid-level to large scale organization.

Both hardware and software support is provided by Microsoft.

Analysing large volumes of data collected from auto ticket barriers at railway stations.

It has allowed fast daily loads and analysis of millions of rows of data, which eventually moved to near real-time.

It handles high volumes of data very well.

It needs more compatibility with common BI tools. 

It does not work well with normal ETL tools. Some functions do not work.

One to three years.