Like my blog? Try my books.
This is the first video I've done on Data Quality Services. It shows how to get started with DQS by creating a new knowledge base and running knowledge discovery on a sample data set.
Data Quality Services, introduced in the previous blog post, includes two main components: DQS Server and DQS Client. There’s also a DQS Cleansing transformation included in SSIS in SQL Server 2012, but I won’t be going into that in this post. In this post, I’ll be talking about these two, core components.
At its simplest, you can think of these two as DQS Server doing all the hard work and DQS Client being the user interface.
DQS Server is installed as part of the SQL Server Setup. It includes database for storing the knowledgebase information and the policies and rules that are applied to business data during cleansing and matching activities. It also includes the DQS engine, which is what applies all the information and rules in the databases.
DQS Client is a stand-alone client that can be used by people in different roles to clean data, maintain the knowledge bases or configure DQS Server. You have to have sufficient permissions, either on the SQL Server as a whole or as a specific DQS role, in order to open the DQS Client and connect to the appropriate DQS Server.
DQS Client allows users and administrators to perform key duties: knowledge base management, data quality projects and administration.
Knowledge Base Management
As described in the previous post, the concept of a knowledge base is core to DQS. Within the knowledge base, you have a number of domains. A domain is a representation of a type of data that is used by the business, such as a customer name, an address, or a contact phone number. For each of these domains, there are certain values which can be trusted and certain values that are invalid. The example I used in the previous post was of counties. County would be a domain, with values like Nottinghamshire and Berkshire as trusted values because they are known to be real counties. You can include synonyms, so that the value of Notts is known to be referring to the same thing as Nottinghamshire.
Within the Knowledge Base Management part of the DQS Client, you can create and manage domains, set the trusted values and give examples of invalid values. Each knowledge base contains one or more managed domain.
Defining your knowledge base and all the domains, values and rules within it can be done manually through the DQS Client. It can also be done through the Knowledge Discovery process to automatically generic rules that can then be approved and included in the knowledge base.
Data Quality Projects
Once you’ve built your knowledge base and created the rules which are applicable to your data, you need to use them. The Data Quality Projects section of the client tool lets you apply your knowledge base for either cleansing of data or data matching.
Cleansing projects apply rules to find data that is incomplete, inaccurate or not in line with your policies. DQS applies the rules from your knowledge base where applicable and can come up with suggestions if there are some values which are not in the domain, but which are similar to domain values. These suggestions allow you to catch typos without having to think of a rule for every single mistake that might possibly be made.
Matching projects pull in the data and look for records which are similar. DQS applies the matching rules defined in the client and, using things like synonyms and similarity of phrases, highlights duplicate records. It can combine records where there is a definite match or highlight ones where there is a possible match, including a confidence score based on the similarity of the records and the specifics of the rules.
Both of these projects then allow you to export the cleaned data so that it can be used by your systems.
Administration
This section of the tool lets you track on-going activities within DQS and update the configuration settings.
A data steward can use this part of the tool to view what processes are in progress within DQS. That person can then choose to stop processes if required.
This is also the place to define the settings for using reference data. The idea of reference data is that you can pull in information from an external source, for example the Azure data marketplace, and use that as a comparison to clean up internal data (e.g. matching addresses to postcodes). You would use the administration part of the DQS Client to configure these connections.
You can also change some settings in this section, for things like logging and notifications.
So you can see that the DQS Client tool is a vital component of a data quality project because this is where so much of DQS functionality is implemented from.
The previous two blog posts have been focusing on Master Data Services. Given the title of this series, it would be remiss of me to forget Data Quality Services, the second piece of the data quality story in SQL Server 2012. Data Quality Services is a new set of functionality in SQL Server 2012, based around the concept of knowledge-driven quality.
The idea behind this concept is that there is a big difference between what’s invalid according to a computer system and what’s invalid according to common sense. Some system might have a field for a person’s age and the system thinks a value is valid if the age is given in numerical form as an integer. The system wouldn’t have a problem with an age listed as 222. A human being looking at this would spot this as an anomaly and guess that the person entering the age mistyped 22. Another example could be counties, stored as string values. A human being would know that Nottinghamshire is a valid county but Narnia isn’t. An automated system wouldn’t spot the mistake.
This is where a knowledge-driven solution comes in. The idea is to take these pieces of knowledge that we think of as common sense and store them in such a way that the business applications and databases can check values against a knowledge base. This knowledge base might include a list of counties so you can have the systems check all values entered in the County field of a database and look for anomalies.
You then have various rules for how these anomalies are treated. You might create automatic rules, for example to change the abbreviation Notts to the full Nottinghamshire so that you can have consistent values without needing a human being to make any changes. You might also have people, known as data stewards, who are alerted to suspected anomalies so that they can check whether values need to be fixed.
A data quality project tends to have two elements to it. One is an initial fix to clean up bad data. This is known as a data cleansing project. As the name implies, the end goal is to have a set of clean data. The tools look through the data, transforming values to match a standard, flagging outlying values that might be anomalies and suggesting changes that could be made. It also hunts for possible duplicates through data matching, applying policies to look for entries in the database that might refer to the same thing. For example, the entries for Jess Meats and Jessica Meats might be flagged up as a possible match. After the cleansing and matching, the output is delivered as a set of data that has been cured of as many inaccuracies as possible.
The second part of a data quality project is what happens next to keep the data clean. As with Master Data Management, this isn’t a fix-once act. It’s very easy for data quality issues to creep back in after the cleansing has taken place so an implementation of Data Quality Services needs to bear in mind what should happen next. The processes and policies need to be defined to ensure that the data quality knowledgebase is used in future to maintain the quality of the data. It’s also important to identify the data stewards who will be responsible for fixing any problems the knowledgebase flags.
It’s also important to think of the knowledgebase as an on-going project. Things change. Data changes. The knowledgebase should therefore also change. The set of knowledge and rules within the knowledgebase can grow over time, bringing more control and accuracy to your data. As more data passing through the knowledgebase, it becomes more tuned to picking out anomalies and better at identifying what the correct value should be.
A Data Quality Services project should include both the plan for how to clean the data initially and how to maintain quality moving forward.
Before either of these can start, however, you need to define what you want your data to look like. A common example is a male/female field. This seems like it should be simple, but one system might store this as male or female, another might have m or f, another might have 0 or 1, yet another might have 0 and 1 again, but with the digits referring to the opposite gender. If you’re merging systems, moving data between systems or pulling together data from multiple systems for reporting, you can have chaos. A key part of a data quality project is working out what you want to be the correct value. Once you’ve done that, you can start applying the rules to change the other values so you end up with consistency across your whole data set.
So when you’re starting to work with Data Quality Services, first take a look at your existing data and decide what you’d like it to look like. Then you can do your data cleansing and data matching to give yourself a clean and accurate set of data to start with. Then you need to hook your knowledgebase into your processes to ensure data quality moving forward.
Part 4: Data Quality Services Client
In the previous blog post, I introduced the concept of Master Data Management as a methodology to improve data quality by creating a “master” version of important information. Master Data Services is Microsoft’s technology for implementing Master Data Management. It is included in SQL Server 2008 R2 Enterprise and SQL Server 2012 Enterprise and Business Intelligence editions. For the sake of this series of posts, I’ll be focusing on the SQL Server 2012 version, since there are quite a few changes.
One of the most fundamental elements of a Master Data Services solution is defining what the master data should be. You have to define what data is important to the company. This information is referred to as entities. For example, you might decide that a customer should be an entity, or a product, or a service offering. Each of these entities then has various attributes, such as contact name, billing address and date of last contact for the customer example. Every individual customer would then be what’s called a member of the entity “Customer.”
A significant part of planning an MDS implementation is working out what these entities are and what are the key attributes associated to the entities. A significant part of the implementation itself is bringing in the members of these entities and making sure that the attributes are properly populated with accurate information.
Master Data Services provides the framework and tools to maintain this data model and these members going forward. There are some core tools that are included in Master Data Services to enable this.
Master Data Manager
This, as the name implies, is for managing master data. This is the primary tool for Master Data Services, used for creating data models, uploading data, creating business rules and more. It’s a web-based tool, updated in SQL Server 2012 with a nice Silverlight interface.
Master Data Services Configuration Manager
Again, the same sums it up. This tool is used for configuring Master Data Services, specifically the databases and web services. You’ll need to have a database to store master data information, whether the source information, a synchronised copy or just a registry (as discussed in the previous post), and this is the tool you’ll use to set that up.
Master Data Services add-in for Excel
This add-in can be used for a lot of the same functions as the web tool, such as managing entities and uploading data. It’s intended to be accessible to end users while allowing more experience master data professionals to handle data in bulk.
There are other tools for developers around deploying data models and interacting with Master Data Services as a web service. Together, these tools form the technical component of a Master Data Management solution. As I mentioned in the previous post, it’s not enough to just focus on the technology part when implementing one of these solutions. It’s vital to processes by which data quality will be maintained and the people who will be doing it. As part of the planning of an MDS implementation, you should consider who among the users will be making use of each of these tools.
Part 3: Introducing Data Quality Services
Data quality is important and data quality mistakes can be expensive. Imagine a situation where a company has two systems, each with their own data set. One system is for CRM, the other is a financial system. Both systems have information about customers. Now an employee needs to send an invoice to a customer. What happens if the two systems have conflicting information about the customer’s address?
Several different things might happen. If they employee notices the inconsistency, he or she might spend some time finding out why and learn that the customer moved a month ago. The employee can then update the incorrect system (or flag it for someone with appropriate permissions to change it) and send out the invoice to the right address. That’s costly in terms of wasted time. If, on the other hand, the employee doesn’t spot the problem, he or she could send an invoice to the wrong address. This could become more costly as someone may start chasing for an unpaid invoice because the customer never received it. In extreme cases, this could escalate to the point where the company is hiring in debt collectors, the customer is getting harassed about not paying a bill they never received, services get discontinued and the customer gets so fed up they go to a competitor.
Hopefully, the root of the problem would get resolved long before the situation reached that extreme but it could still build up costs along the way. At the very least, there would be delays in the payment of an invoice if it went first to the wrong address.
Another example of costs is if there is duplication. I might be listed in someone’s customer database as both Jess Meats and Jessica Meats. If that company sends out marketing fliers to all of their customers, I would get two. That might not seem like much of a problem until you think about the likelihood of this problem happening with multiple records in the systems.
There are all sorts of ways that data quality issues can waste time or add costs to businesses.
Improving the accuracy of data is one area that Microsoft have been working on with SQL Server 2012. There are two toolsets that are used for this end: Master Data Services and Data Quality Services. In this blog post, I’m focusing the first of these.
Master Data Services is a set of functionality that was introduced to SQL Server in the 2008 R2 release and improved in SQL Server 2012. MDS is Microsoft’s tool for Master Data Management. Master Data Management, or MDM, is all about creating a single true version of data. The concept is that you have a set of data, known as the master data, which is up to date and accurate.
There are two main ways that MDM is used. One is for business intelligence. Essentially, data is pulled out of the various systems in the company, cleaned up, checked for consistency, and then published in its accurate form into a data warehouse which can then be used for analysis and reporting. In this usage, MDM is all about making sure that reports are accurate and that the conclusions drawn from reporting can therefore be relied upon.
The second main form of MDM is operation MDM. In this scenario, MDM is used to ensure that data in all the systems is consistent and clean. For example, if someone updates a customer’s address in the CRM system, processes and procedures ensure that the address is also updated in the finance system, avoiding the scenario described at the start.
There are multiple ways to implement MDM, the three main versions being: transactional hub, registry and hybrid.
In the transactional hub model, you essentially replace the content databases of various systems with the master data database. The company systems write data directly to a central database which is then read by all the others. In some ways, this is the ideal MDM scenario because it ensures that all the systems are literally looking at the same source data. In reality, the technical implementation is incredibly difficult because different systems all treat data in different ways and expect it to be stored in a certain manner.
The registry model leaves the data where it is in the source systems but creates a database pointing to the data that should be viewed. In this case, the master data database doesn’t actually include the data, but instead has pointers to the correct information, allowing data to be reported on accurately. The drawback of this method is that it is read-only – information can be queried via the master data registry but not updated.
The third model is the hybrid model. In this model, data remains in the content databases of the source systems, as with the registry model, but you also have a central repository of data, as in the transactional hub model. In this case, the hub contains a cleaned store of the data, synchronised with the original source systems.
One of the key things to note is that MDM isn’t a fix-once solution. When implementing Master Data Management, whether in SQL Server or using another technology, it’s important to remember that data doesn’t stay fixed. These quality issues come into the systems as changes occur so it’s important to build in procedures to maintain the quality of the data going forward. This can be done by a mixture or automatic procedures and notifying users of changes. An MDM implementation should involve the identifying of data stewards, individuals whose job it should be to maintain certain pieces of the data. For example, a customer account manager should be responsible for maintaining that customer’s contact information, so when something changes, they will be the one to approve changes via MDM to all the systems that include that data. It’s vital that the planning of an MDM solution should include policies for how data quality will be maintained going forward.
Part 2: Introducing Master Data Services
Someone asked this question on a discussion list and I thought the answer was one that others might be interested in.
What do you do if you want to set a field in an InfoPath form to show the month that the form is filled out?
This can be done by using the today() function. This function returns the date of the day that the function is called on. For example, if you call it on the 1st of March, it will return the value 2012-03-01 (you can change the way this is displayed, but this is the value pulled back by the function).
If you want to just display the month, rather than the full date, you can expand the formula using the substring function. This is a little function that takes a string and returns just a piece of it. So the function substring(today(), 6, 2) would return two characters of the string “2012-03-01”, starting with the sixth character. In this example, you’d get 03.
So that gives you the value for the month, but what if you want to see the text?
This requires another rule or, more accurately, twelve of them. Have a rule that looks at the value brought back by the previous formula and if it equals “01”, set the value of the month field to “January” and so on for the other eleven months.
It’s probably easier to keep these two pieces slightly separate, rather than trying to do everything in one set. For example, you can have a field called month_value which is not displayed on the form. Create the first rule to populate this with the value substring(today(), 6, 2). This rule could be attached to form load behaviour, a button, another field, the submit action or various other items. However you choose to trigger it, it will populate this field with the number of the month.
Then you should have a second field, which is used to display the month name in the form. On this field, add the twelve rules to set the value of the field based on the value in month_value. You may want to set another condition that only does this if the month field is blank.
And make sure to make this field read-only otherwise someone will type something in and wreck your lovely rules.
Myself and colleagues (mostly Sam Solieman, but also will Thompson) have been running a series of webcasts on SQL Server 2012. The ones we've run so far have been recorded, so you can watch them on demand. The recordings are on the Partner Learning Center, so you'll need a Microsoft partner login to access them.
The sessions that have been run and recorded so far in this series are:
The next webcast in the series is on Thursday 2nd February from 4pm. Sam will be talking about the security enhancements in SQL Server 2012.
Still to cover are:
To register for any of the upcoming sessions, go here.
The SQL Server team blog has posted recently about customers using SQL Server 2012 today. This post has five case studies of big customers who are on the early adopters programs for SQL Server, putting to use the new features of SQL Server 2012.
This nice video has people giving their comments on their favourite features of SQL Server 2012.
I'm going to be speaking at the International SharePoint Conference in London on 23rd to 25th April. My subject is: No Budget, No Time, No Permissions. It's a case study of building a SharePoint site with no allocated resources. Come along to find out more...
Registration for the conference is open now.
http://mssalesdemos.com now includes a demo machine for SQL Server 2012. This includes the Telco demo that was present in the BI X demo machine as well as new demos on SQL Server 2012 functionality. There's a set of demos on PowerPivot, including the auto sales demo which has been shown at various events and one on the food consumed on picnics. There's also a demo on the new Master Data Services functionality.
These demos have a distinctly BI flavour but the machine is configurable if you want to build some demos of your own.
This machine is available either hosted or as a download. Just be prepared to wait a long time for the download.
If you're a Microsoft partner and you've tried to find training courses and online learning offered by us, you'll probably be familiar with the joys of the Partner Learning Center. The Partner Learning Center stores information on what sometimes feels like every training course ever offered to partners. It's therefore a wonderful resource. It's also, for obvious reasons, very tricky to figure out where to start. If you search for a common product name, such as SharePoint or SQL Server, you could get loads of results. Yes, you can filter the results so you're not looking at classroom courses taking place a thousand miles away, but it's still difficult to know where to start if you're learning a technology from scratch.
Introducing: Microsoft Learning Plans.
When you log into this website, you can build learning plans based on various factors. If you're trying to gain a competency, you can get a learning plan of technical training relating to that competency. Or you can choose a learning plan based on your role or a particular technology you're interested in.
The end result is a list of available resources and online courses (and links to exams and/or classroom courses). A lot of the material is actually on the Partner Learning Center, but this way you know where to start and can work through the available material in a sensible order.
If you're getting started with a piece of Microsoft technology, this website is a very valuable resource.
I've been playing around with Power View (formerly codename Project Crescent). I started with a simple PowerPivot file I'd created to analyse the transactions of a small dog grooming salon (I wanted something different from the usual Contoso demos). It was a simple matter to publish this PowerPivot file to a PowerPivot gallery in SharePoint 2010.
With the file saved to the gallery, I could click on the new icon to create my report (my screenshot still calls it Crescent, since I'm using CTP3 code).
This gave me a blank canvas from which to start creating my interactive report. I could quickly and easily create rich visualisations, including charts which move based on a timeline, with just a few clicks.
It's also possible to save this visualisations as various views, making it simple to jump to different data representations to present conclusions from the analysis.
I'm organising an event in Microsoft's London office on 5th December. The name of the event is SharePoint: The Bigger Picture and the aim is to talk to partners who do some work with SharePoint but who aren't taking advantage of all the capabilities. We'll be talking about the range of high-end capabilities SharePoint can provide and the services opportunities involved for partners. The event should be an eye-opener to the potential value SharePoint can provide Microsoft partners.
We'll have in attendence representatives of SharePoint training partner Combined Knowledge in case attendees want to get skilled up in new areas. We'll also have a couple of established Gold partners speaking in case the attendee partners would prefer to go down a partnership route.
The event will be focused more on the business than the technical and will assume some prior experience with SharePoint.
If you're interested in finding out more, the registration details are here.
These guys at SQL PASS presented on using Kinect to interact with SQL. Create databases by raising your arms. Delete databases by kicking them away. Back up databases by bringing your arms back.
Awesome!
SQL Server 2008 R2 comes with a lot of features and functionality. Even the Standard version has a long list of capabilities to set DBAs drooling. But then there’s the Enterprise version. A question that comes up, more often than we’d like, is, “Why would I need Enterprise?” If there are so many features available out of the box with SQL Server Standard, why go to the Enterprise version?
If you want to go into it feature-by-feature, the SQL Server website has a comparison site which allows you to look at the features under various headings and see which versions of the product those features are included in.
But if you don’t want to go through the whole list, here are my thoughts on some key ones.
Master Data Services
As a set of features, Master Data Services comes with SQL Server Enterprise but not with the Standard edition. MDS provides the technical framework of a master data management solution, letting organisations have control over their data. MDS helps companies ensure that their data is consistent, accurate and up to date, even if that data is spread across multiple applications and databases.
As a concept, master data management is a very powerful one. There are few companies who would say that they don’t want to be confident that their data is correct. If you’re building a business intelligence solution, you want to be sure that the data you’re starting from is valid. The best analysis in the world won’t give you reliable answers if you give it figures that are wildly incorrect. MDS can provide a master data database to store an authoritative source of information to be used in applications. It can also provide a framework of checks and rules to ensure that data stays consistent and correct across multiple applications if values are changed in one.
PowerPivot in SharePoint
PowerPivot starts off as a free download add-in for Excel. Users running Excel 2010 can install the client add-in and start creating powerful dashboards and applications using familiar pivot table interfaces that pull data from a range of sources, both on-premise and in the cloud.
When users want to share these PowerPivot files, that’s where the SharePoint add-in comes in. This lets users publish PowerPivot files to galleries in SharePoint where they can be viewed by others in the organisation. It also provides a management dashboard, so IT can control permissions to the files, audit changes and manage the refresh schedule for the data. This allows IT to have control over files produced quickly by end users.
The PowerPivot operational dashboard and SharePoint collaboration layer come with the Enterprise edition of SQL Server.
Auditing
SQL Server Enterprise allows you to set up automatic auditing of changes on your databases. You can have fine grained control over what events get logged. Depending on your data and application, auditing might not be optional. Some industries have regulations regarding data meaning that you need to be able to show who has accessed and amended values in a database. If that’s the case for you, the Enterprise edition is almost essential.
High Availability
For some databases and systems, downtime is not an option. Sometimes, an application going down means that money isn’t coming in. For those applications, a high-availability strategy is needed. SQL Server Standard includes some features to enable high availability, but the Enterprise edition brings more. Features like database snapshots, mirrored backups and fast restore mean that when something goes wrong, a back-up system can be online and in the right state to carry on.
Scalability
SQL Server Enterprise can scale more than the Standard edition. In hard numbers, Enterprise can run on 8 CPUs instead of Standard’s 4, utilising 2TB of memory instead of 64GB. The maximum database size is the same (524PB) for both, but the limits on CPUs and memory will affect the performance of a big SQL deployment. Some of the options around indexing will also make a big difference on performance when the database size and usage gets large.
These are such some examples of areas where SQL Server Enterprise provides a significant advantage over SQL Server Standard. For the full list, check here.
Like a lot of people around the world, I've been watching the Build Keynote. During this opening of the Build Conference, Steven Sinofsky presented about the upcoming Windows 8 and there were a lot of demos. The first lot of demos focused on the new Metro UI, which seems geared particularly for slate devices. It resembles the phone somewhat and is slick and shiny.
They showed a few different styles of devices, from ultra-light netbooks to massively powerful gamer machines, along with demos of amazingly fast start ups. (You'll probably get a better feel for the keynote if you stick the word "super" into that sentence a few times: "super fast", "super light", "super powerful", "super super.")
One thing I was concerned about, while watching the new UI demos, was that it seemed more geared to a fun slate than a professional computer. I was relieved to see some demos later on addressing that, showing the keyboard and mouse experience, multi-screen experience and some enterprise features such as security.
The keynote is a couple of hours long and packed full of demos so you can really only take in a fraction of the new features that get shown. Some look really useful. Some are shiny gimicks. Some are geared towards developers. Some will be important to everyone.
I definitely suggest watching the keynote and getting a feel for what's coming down the road from Windows.
SQL Server codename Denali is in beta and the third preview has been available for a little while, which is great if you have the time and inclination to install, configure and run Denali for yourself. If you want to make use of a demo machine that someone else has already configured, there's one now available on Microsoft Downloads. This download is available in a lot of parts. It's a big machine but if you want to start playing with the next version of SQL Server, you now can.
Registration is open for the 9th SQL Bits, this time being held in Liverpool. The conference runs from 29th September to 1st October and includes some top speakers talking all things SQL Server.
Find out more or register here: http://www.sqlbits.com/
There is a new Product Guide available for SQL Server Codename Denali, the next version of SQL Server. This product guide includes PowerPoint decks, technical whitepapers, datasheets, demonstrations and more.
If you want to see what's coming in the next version of SQL, take a look.
There is a new training offering for partners on Business Intelligence. The Practice Accelerator is a technical learning offering, delivered in a series of webinars over 4 days, giving 16 hours of online training in total. As well as this, the offering includes documentation, guides, hands on lab and other materials that can be used after the training session to deliver strong BI projects.
Delivery dates for this new offering are 25th July - 28th July and 19th September - 22nd September.
This is delivered by the PTS team and comes out of a partners Technical Advisory Services hours. These hours of technical support come as part of the partner benefits for Silver and Gold Partners. The Practice Accelerator costs 10 hours and allows up to four consultants to take part.
Link to register is here:
https://training.partner.microsoft.com/learning/app/management/registrationex/LMS_Registration.aspx?UserMode=0&Mode=0&ActivityID=746096
BI is a big opportunity for partners. Gartner recently released analysis of the BI market and found that spending on BI and analytics software had increased by 13.4 per cent in 2010. This is a huge jump, particularly considering the tight budgets worldwide at the moment.
Maybe the recession has played a part in the demand for BI. Companies need to be able to investigate their own operations and get a clear picture of what is working well and what isn’t. If a company is having to make tough calls about where they should invest their money, they need accurate information to base their decisions on. A good BI implementation can provide them with that insight into their numbers.
In the BI space, there are four major players fighting for this growth share. Microsoft is up there with SAP, Oracle and IBM. And remember, back in February, Gartner put Microsoft at the top of their magic quadrant in terms of ability to execute in BI.
This isn’t nearly as common a myth as some of the others I’ve described so far in this series. Compared to most of the big players in this space, SQL Server is very low priced. The list prices for our competitors in this space tend to be about 10x that of SQL. The difficulty comes when those other companies offer massive license discounts to try and get customers to go with their offering.
When that happens, it’s worth doing the sums. There are other costs to consider beyond just the licenses. There are support costs to consider. Some companies make more money from support contracts than they do from licenses. In those cases, that’s where your wallet will be hit. You should also think about the number of DBAs you’ll need and the costs associated with them. Then there’s the cost of integration.
Microsoft’s solutions tend to be built alongside one another, and designed with integration in mind. So if you have other Microsoft products, your integration headache could be substantially reduced.
But then there’s always open source. Some customers make the mistake of believing that open source equals free. For many open source providers, it’s true that there aren’t license costs. But that’s not true for all of the open source companies and you have to remember that all businesses have to make money somewhere. Look at the costs of support and services and you may find that the costs are much higher than you anticipated.
In the most recent magic quadrant from Gartner about Data Warehouse Database Management, they said: “[Microsoft’s] license cost profile is comparable to or less than open source BI vendors and considerably less than its commercial competitors.”
So the analysts think that SQL is comparable in cost even to open source providers. Then there’s the fact that SQL Server set a world record for price/performance in databases.
Myth: busted!
If you want to know more about the SQL Server Myths, I'm running a short sales training webinar on June 6th at 4pm GMT. You can register here.
In terms of big customers, Microsoft have been playing catch up with regard to databases. Some of our competitors have had a much stronger foothold into the enterprise space, with huge deployments in huge customers. That’s probably where this myth originates. But the fact is that we’ve been closing the gap for some years now.
A 2009 report on database market share showed Microsoft in second place to Oracle. One key thing to note though is that the rankings were determined by revenue. Given the differences in the cost of SQL licenses and our competitors, is it surprising that Microsoft would be behind, particularly taking into account the huge discounts offered to charitable customers? If you look at the same data in terms of units shipped, from that same report, Microsoft has shipped more units than IBM and Oracle combined.
Gartner’s latest magic quadrant put Microsoft firmly in the leaders quadrant, above all the others in terms of ability to execute.
This is a myth that Microsoft have been fighting against for years. Some bad history has left us with a reputation that we’ve been struggling to overcome ever since. To get over this history, Microsoft have come up with procedures around security, fault-discovery and patching that are incredibly tight, with many companies aspiring to copy our techniques.
Microsoft are leading the way in terms of time computers are vulnerable. This is measured as being the time between a vulnerability being discovered and it being fixed.
In terms of numbers of flaws, Microsoft are ahead of the other players, particularly in terms of the number of high severity vulnerabilities. And it’s not just Microsoft saying this. When Forrester wrote about SQL Server 2008, they said: “Unlike Oracle and IBM, SQL Server continues to enjoy the lowest database security vulnerabilities in the industry.”
That was three years ago and we’ve continued this trend.