SQL Server Reviews

• SQL Server Integration Services tools

• Data Quality Services tools

• Master Data Services tools

• SQL Server Reporting Services tools

• Data Partitioning (Horizontal Partitioning) tools

• ColumnStore Indices

• In-Memory OLTP tables

We design and implement DW solutions with SS 2016 Dev Ed tools. This solution has given us high levels of productivity, which has allowed us to use an Agile approach to the design and implementation of DW solutions for our customers, and this Agile approach has in turn given us a competitive advantage in our market.

We have also started exploring the use of Microsoft R Client, MS R Services and MS R Server with SQL Server 2016 Dev Ed, which are part of the hosted Data Science package.

We are also looking forward the inclusion of Python support in SQL Server 2017 for said hosted Data Science package.

We are very interested in complementing DW solutions with Data Science and Machine Learning solutions, which could be a major plus for our existing DW customers, even though all hosted data science tools are only available in Enterprise Ed (for our customers), which again presents the same limiting factor (budget) already mentioned.

Having said that, we see that the potential that the hosted Data Science tools offer to some of our customers is large enough to be explored and considered on a case by case basis, with proper ROI analysis.

We do not have much to complain about SS 2016 Dev Ed in itself, we do have some complaints regarding licensing for SQL Server 2016 Enterprise Ed. In an emerging market like Argentina, it is very steep for our customers to pay U$S 28,000 or more on licensing for an instance of SS 2016 Enterprise Ed, and this poses a limiting factor to our growth.

We have been using SS 2016 Dev Ed for a few months so far, but have been using the previous version (SS 2014 Dev Ed) for more than two years.

The deployment tools for DW solutions in both 2014 and 2016 editions of SQL Server are part of the SQL Server Data Tools (SSDT) package. As SSDT is based on Visual Studio, this toolset has been very stable since its inception, both in terms of performance as well as in terms of functionality, so, deployment in 2016 is done in the same way as in 2014, which translates into no issues during deployment.

There were no stability issues.

There were no scalability issues, even though the Dev Ed does not offer the same level of performance and scalability that the equivalent Enterprise Ed offers.

Pertaining to the hosted Data Science package that we have been exploring, we have found an important increase in scalability when comparing the performance of a given solution running as hosted, and the same solution running on the same server with only client R tools.

This scalability advantage presents itself as an important reason to consider these tools as a viable solution to some of our DW customers.

As usual, MS offers a very good customer service. The amount of resources (self-study materials, online courses, tutorials) is huge, most of it is free.

Paid customer service is also very good.

Paid tech support is very good and efficient.

I have always used DW tools from Microsoft since SQL Server 2000.

If prior to the setup you do the proper training, there are no issues with setup, but the learning curve is wide and tall.

You could get started fast and sure if you stick to the many Wizards included with the tools, but the scope of said Wizards is limited.

We did not have deployment/implementation issues.

Since SS 2014 Dev Ed, MS is offering these tools free of charge, ROI mainly is focused around training investment. As I have said, we do the training in-house. ROI is around one year (12 months).

No issues with pricing and licensing for SS 2016 Dev Ed, as it is free of charge, as mentioned above, the thorny issue with pricing and licensing is with customers. We do our best to design DW solutions that can cover reqs from our customers within the capabilities of SS 2016 Standard Ed.

The cost-benefit ratio offered by Microsoft's DW solutions is, by a long shot, much more convenient for our customers (small and medium companies) than solutions from other vendors, hands down.

Consider the ROI (most training investment). If training is not in-house, only hire training from an official Microsoft Training Center in your region. Look for the best Training Center. Once you are done with the training, you can start taking customers for DW projects.

Most of SQL Server High Availability and Disaster Recovery solutions. They are easy to configure and maintain.

It provides the best performance and an easy way to monitor and troubleshoot problems.

Indexing, execution plans, and the SQL Server Management Studio performance.

Over seven years.

No.

No.

Medium.

No.

Complex design, easy configuration.

It is not very expensive and is suitable for an international company, like what I am working with. Free licenses are suitable for small companies, too.

No.

Stable and easy to administrate.

Performance improvements, optimizer enhancements. Most of our clients have high demands for performance, and this version of SQL Server delivers what we need.

We are a professional services company, so we use SQL Server to help our clients achieve their goals. Our clients use SQL Server 2016 for their most demanding mission critical systems, for data warehouses, and big data solutions.

The Query Store is a good start, but I expect the query processor to be a lot smarter and to use machine learning in order to improve and adjust execution plans automatically.

One year.

Since we are working with a lot of clients on many edge cases, we encounter bugs and stability issues once in a while, but these are rare.

If you know how to work with the product and leverage its various features and possibilities, then you can achieve great scalability.

Microsoft offers several levels of technical support, which is OK, but not too good. But there is a wonderful community with lots of resources on the internet, so most issues can be solved without contacting Microsoft support.

No, I have been using SQL Server for the past 20 years.

The initial setup is very clear and friendly. It has improved from the previous version.

Unfortunately, SQL Server licensing is a very complex topic. I advise people to consult with a licensing expert.

No.

Download the Developer Edition for free, install it on your personal computer (it’s very easy), and start exploring. If you need help with something, just search for it on the internet, and you’ll find a wealth of resources about everything you need.

• Net integration is best for me
• Better optimizer
• Ease of use (admin tools)

We used this to solve many complex problems.

Overall, trying to make over with Oracle, it can be improved in usage for high data migration companies of large scale.

I have experience using both the versions of SQL Server 2008, 2012 for more than two years.

Never had issues with stability, may be we were handling structured data well.

No, federal vs clusters approach explains you well that you can search for them.

Technical support rating would be around a seven and a half out of 10.

Setup is easy as what we want the database to do for our problems, compared to Oracle latest versions you need to spend more time on the initial setting up where these cannot go wrong with partitioning and indexing and it needs to be perfect.

Go for SQL and compare it with pricing and stability with other Microsoft products.

Definitely yes, with Oracle only if the needed situations cannot be met, mostly it covers better.

Best practices are the always a better choice for implementing the solution, hire a good expert and consult a better person with huge experience in architecting database systems for decades.

The way the product shows the execution plan and the facility to the developer and tests the SQL code.

The product is used in the core systems. It has been used since the beginning.

Better execution plans and better debugging.

Improved dependency of use of temporary tables to reach good performance in complex SQL.

I've used this solution for six years.

Sometimes the cross-database performance isn't the best.

No.

An eight out of 10.

I have used Oracle and I also use IBM DB2. All of them have pros and cons.

No.

Great product. Like any software, a good and prepared staff is the key to success.

The ability to organize and structure our data in order to extract and provide it to various products and systems within our organization. It is widely used and has great online support from many third-party sites.

We're able to easily upscale previous systems to provide an efficient interface for our internal clients to do their jobs effectively.

Learning SQL could be easier. Some inbuilt tools to enable faster query writing would be very useful.

I have used various versions for at least 25 years, on and off.

No, none.

When upsizing, some tools do not do a very good job of scaling efficiently and lots of after tweaks need to be made.

From Microsoft -- poor/very expensive, but there are many other helpful resources out there.

Used MS Access, but data ports became too large for it to support, so we needed a product to take us to the next level as our data ports grew.

A lot of initial reading was required to carefully manage the process of setting everything up correctly.

Not my area.

No, others evaluated.

Do your homework first and be prepared for a tough learning curve.

Probably the most useful feature of SQL Server is the ability to write and execute SQL on the fly.

Even though there are some very useful features on Oracle, MySQL, and other platforms, the SQL Server by far has the most robust capabilities of any database platform.

In my experience with numerous coding languages and platforms, the SQL Server has the only programming language that allows the user to create, compile, and execute code in its own language.

To clarify, Java, .NET, PL/SQL and all other programming languages can dynamically create code, but not their own. In other words, Java can dynamically create SQL and execute it, but it cannot create Java and compile/execute.

Other great features are:

• Passing tables as parameters
• Table valued functions
• Horizontal table partitioning
• Very granular disk partitioning

The most recent example is a data warehouse I've created for a client that enables us to use a "no-SQL" construct. This is only possible due to the dynamic SQL capability.

Our client collects data from dozens of sources with little to no commonality between them. With other platforms, this would require a table for each data source. However, because of the dynamic SQL, we have three tables that will accommodate ANY data source and it will never require us to change the data warehouse schema.

As a result, maintenance is virtually zero.

The only real improvement I've been looking for is finally being addressed by Microsoft.

Since SQL Server only ran on Windows, it was not competitive with other platforms which could run on Linux. This has recently been realized with the release of SQL Server for Linux. I currently have the pre-release version and I'm very impressed with what they have so far.

I have been using SQL Server for 17 years.

We did not encounter any issues with stability. None at all.

We did not encounter any issues with scalability. I have been able to create databases with billions of records with no degradation in performance. The partitioning has been a critical feature in enabling scalability.

In my experience with their support, I would rate it as outstanding. Their techs are professional and extremely helpful.

I typically use whatever database platform my client uses. However, whenever I am provided with the option to choose, I will always go with SQL Server.

In older versions, the setup was rather onerous. However, in the latest several releases, it has been extremely simple to install and set up.

Do the research and get the correct licensing model for your given purpose. A lot of people gravitate toward the Open Source databases because they don't have an upfront cost.

I find that what you don't pay upfront is what you have to invest in development and maintenance time on implementation. On far too many occasions, I have spent weeks writing code for features that SQL Server already has built in.

Either pay for the licensing cost or pay multiple times that for the labor involved in creating features, from scratch, that are native to products like SQL Server and Oracle.

I have a good deal of knowledge on SQL Server, and Oracle, MySQL. I have some familiarity with DB2 and PostgreSQL. The database platform chosen will ultimately depend on the needs of the client.

Look very closely at the built-in features. For those features that you may need, estimate what it would take to replicate that same functionality on the "free" products.

The comparison is not on the licensing cost. It's on the features and the license cost versus the labor cost.

The most valuable feature is the easy installation. The user just needs to know how to read.

The SQL Server is helping us with the most feared corporate problem: Excel spreadsheets. The individual user can perform an integration and share the same results with others without the necessity to "share" the file. This reduces the data traffic on my company network.

Table partition and memory management. The performance could be better.

I have used it for more than 10 years.

The SQL Server has a big issue when some systems use Java as a primary software interface. This database loses the ability to manage memory, consequently locking the processes and losing performance in the execution of some robust queries.

Unfortunately, this database doesn't have a good concept of partition table. If you need to create monthly (Jan. to Dec.) partitions in one table, this action creates 12 different files for the same table. In others databases, this process is more transparent and capable to create partitions inside the same database.

The technical support team for the SQL Server is very friendly, if we compare with others commercial database products. Not only with Microsoft, but this database has more technical information published on the internet, books, and self-taught users to help.

I am still using other commercial solutions, but the price of this database is much less expensive than others. It is about four times less expensive.

The initial setup is one of the good things about MS SQL. It is very easy to do and start the development that you need to use after the installation.

In my point of view, the MS SQL is the most inexpensive database commercial solution. If do you need to build a consistent ERP for example, with a medium to hard capability, and don't have much budget to spend it, this is your solution.

Of course, I evaluated other options! I never choose the first option and I always try to look around to find competitive vendor options. I already know and have expertise with IBM/DB2 and Oracle solutions. For this new architecture, MS SQL was the best option at the moment for being a pilot project.

I suggest that you are first familiar with the bundle functions and plan some of the functions before starting the project. Find more information about the routines and how easy or hard it might be to start the development when you are thinking about cascade and constraints.

There are many, but I would have to say the two most important for me have been SQL Server Profiler -- to run traces and the ability to kill sessions in the enterprise manager tool. This command-based functionality lets you search and sort through processes actively running on the system to find your i/o hog – that is the user that is overloading the database either through a hung query or bad SQL code.

Once the hog user is found, the terminate or “kill” command can be used to shut down the hanging transaction. Similar to how clearing a vehicle crash on the interstate frees up waiting traffic, this kill command gives other users the ability to continue work as normal once the terminated transaction has been killed. This should be used sparingly though since it can cause waiting applications to corrupt data depending on how the application was written, but it is an often used research feature when a system starts to overload or show major signs of slowing.

In an effort to forego a data warehouse purchase, one company I worked with used a second SQL Server instance loaded to separate hardware as a reporting environment avoiding the need to hire additional employees to support a data warehouse. Essentially, the production database was snapshot and copied nightly to the reporting instance where it was restored through automated processes.

All employees who wrote queries and reports against the reporting instance utilized the same knowledge, skillset and tools already used in the OLTP production environment. While it didn’t have all the abilities of a true data warehouse, it was quickly implemented and well served for the desired purpose.

Multiple operating systems support should improve. The ongoing and newly marketed support for Linux and Unix environments running SQL Server is a big win for Microsoft, in my opinion.

Previous hold back related to open source environments stemmed from admins or management who would not run SQL Server for various reasons which many times boiled down to not liking Microsoft as a company.

Open source and Microsoft have conflicted in the open source world for years, but I think Microsoft is finally starting to turn that tide in an effort to compete with other vendors.

20 years. I began as a Microsoft Access database programmer in 1998 connecting early dynamic intranet websites to Access backend databases.

I later progressed to the SQL Server 2000 environment and have utilized SQL Server ever since then.

Issues encountered with stability were always correctable assuming we as a business were willing to spend money. The two major issues that come to mind were slowing of image files loading to disk using SQL Server and lack of proper indexing.

Images were archived for a 30 year legal period at a rate of at least 12 million per year and we eventually just hit a max depth of data where SQL Server had a hard time returning results. Shortening the physical file path structure of how deep SQL Server had to dig through the hard drive in addition to changing to new and improved disk technologies resolved the problem.

After this drive change occurred, a separate problem started where our database came to a crawl which is when we realized our regularly scheduled database index had not been updated to accommodate the new structure changes. A new index was put in place which is when SQL Server started humming along better than ever.

Microsoft makes life easy to expand your environment through clustering tools and after-hours operational scheduling features. SQL Server is easily suited for small businesses where it found huge popularity, but daily operations can sometimes be overlooked as the business grows causing intermittent operational kinks. I have seen a few small businesses determine that a slowing or overloaded SQL Server environment means they should look at more enterprise level databases, which is simply not the case.

Just like a Windows PC needs basic defragmentation maintenance to run effectively, SQL Server needs its own regular maintenance. There are many options available to improve the performance of SQL Server including the simple add-on clustering features which will balance excess load on the server.

In addition, database indexes need to either be created or re-indexed periodically. A database does not automatically know how an application was designed although it’s very good at trying to guess using built-in scheme mapping software. Creating indexes and related maintenance schedules specific to your environment can make a huge difference in how quickly SQL Server responds to data requests.

Disk partitions are another method for improvement. No matter how well SQL Server software is configured, it is limited by the hardware level. Underlying disk usage grows as data grows meaning the more data you add to the database results in the longer it will take for the database to find data on-disk. That’s when it’s time to either spread data across many independent disk sets or move to more expensive flash drives which save time by avoiding read & write disk operations.

These options, of course, require time, effort, and money, but they have been well worth the costs of doing business based on my experience.

I have only had to call Microsoft support 5 times in my nearly 20-year career, but they were extremely helpful. The one frustrating experience was 10 years ago when I worked over the phone 8 hours straight with 3 separate techs from the India location. Keep in mind, these were days pre-remote support meaning I had to verbally translate every diagnostic error I was seeing on the screen and assume the tech was clear on the meaning. The techs I worked with were extremely nice and tried hard to help, but I reached my frustration level due to the verbal language barrier.

The senior tech, who I spent the most time working with, had an extremely thick accent and was hard to understand. I could tell he had hit a dead-end on his knowledge of the product so I finally asked to be transferred to a specialist in America. Once I found the right person in the US, my problem was resolved within an hour.

In hindsight, I knew the senior tech was lost at the 6-hour mark, so I should have spoken up then, but he never complained and kept trying different solutions, which is good on Microsoft’s part.

My team and I have tried many different technologies including MySQL (cheap but effective), Oracle (expensive and effective), PostgreSQL, and DB2. I never jumped on the PostgreSQL or DB2 train, but could not give you a specific reason why due to my limited knowledge of the products. It usually came down to lack of knowledge for available programmers in our area, meaning we would have to train new hires and take a lot of time getting them familiar with a new database structure. That defaulted us to either Oracle or SQL Server since MySQL was not used in production at the time due to limitations surrounding support.

Oracle owns MySQL these days and they, of course, would prefer you run full fledged Oracle database for support needs. Oracle’s supremely expensive licensing has normally pushed me and coworkers to Microsoft SQL Server although every organization I have worked with pays for some form of Oracle even though SQL Server is primarily touching end users.

I personally feel Oracle is a great database but also think Microsoft SQL Server can be configured to run just as well as Oracle in most cases. The problem I normally find is that many bloated applications run SQL Server where more streamlined (and many times less functional) applications run Oracle. For that reason primarily, Oracle has had a better reputation in the pre-Amazon world.

This could all change in the years to come as Microsoft starts to fight Amazon and Google in cloud processing.

Creating a SQL Server instance using the pre-installed GUI setup screen available in either Visual Studio or Enterprise Manager can be very simple to create and maintain database “instances” making it a prime choice for small startup businesses.

In addition, Microsoft has added loads of training videos on their website along with step-by-step instructions for creating and maintaining servers. The user can get as complex as desired in SQL Server by learning all the behind-the-scenes commands the GUI is using. You will start researching commands quickly once the first SQL Server error occurs and the GUI doesn’t know how to resolve it.

For the most part, what you see is what you get with Microsoft’s licensing website. I sat through many hours of negotiations with many database and application vendors. Some application vendors try negotiating a better deal based on their licensed volume, but only twice have I seen Microsoft cater to lower licensing and it was because they wanted to make long-term wins with the customer knowing they had the possibility of gaining market share.

That said, Microsoft is still usually cheaper than Oracle who will sometimes look cheaper to start with but add the possibility of higher long term rates. Asking for a discount can never hurt.

MySQL (now owned by Oracle) is sometimes used internally for dynamic website needs, but Oracle is the only competitor evaluated for SQL Server.

Start in the cloud if feasibly possible and if it makes sense for your business. I have mostly worked with organizations that still don’t trust the cloud for security and legal reasons, but Amazon, Microsoft, and Google are releasing promising products in the cloud that are leaps and bounds ahead of processing power for local servers assuming your organization is willing to pay what it takes. I have priced cloud services enough to know they’re expensive, but they could replace a lot of unknowns for growing businesses or those starting from scratch; mainly network security, redundancy and technical skillset of employees.

Cloud services have full-time employees focusing on those niches meaning you as an employer will have less headaches at night.

According to CNBC and Synergy Research Group as of this writing, Amazon AWS leads 33% of global cloud market share. That may sound like a small number, but it’s by far the majority of customers since Amazon’s AWS cloud revenue is more than the next five providers combined. Two of those 5 providers are Microsoft and Google.

And for those of you new to the cloud, you do have the ability to run Microsoft products, including Microsoft SQL Server, in the AWS cloud.

Database Engine

Integration Services

Reporting Services

I am a Senior Software Engineer. I've used SQL Server for most of the solutions that I worked with, from web crawlers to logistic platforms and medical devices.

• The Server Objects area of SQL Server Management Studio
• The SQL Server Configuration Manager

Over 10 years.

Yes, I encountered issues only when using the Express editions during the development. SQL Server, generally speaking, is stable also in high transaction environments.

Yes, I encountered issues in the CPU intensive application, like web crawlers. It was difficult to reach the desired scalability through the fine tuning options available.

The support is good. The knowledge base is very good.

I have experience with Sybase SQL Anywhere .

I switched mainly because the SQL Server is widely deployed and known, for better XML support, and for better integration with .NET framework and its related technologies.

It was straightforward for most of the installation, thanks to the wizards.

I am not a direct purchaser, so I don’t know.

SAP/Sybase SQL Anywhere for the small footprint which makes it suitable for embedded solutions.

SAP/Sybase SQL Anywhere for the built-in synchronization technologies available.

Improve the built-in database replication technologies.