February 5, 2011

Comments on the Gartner 2010/2011 Data Warehouse Database Management Systems Magic Quadrant

Edit: Comments on the February, 2012 Gartner Magic Quadrant for Data Warehouse Database Management Systems — and on the companies reviewed in it — are now up.

The Gartner 2010 Data Warehouse Database Management Systems Magic Quadrant is out. I shall now comment, just as I did to varying degrees on the 2009, 2008, 2007, and 2006 Gartner Data Warehouse Database Management System Magic Quadrants.

Note: Links to Gartner Magic Quadrants tend to be unstable. Please alert me if any problems arise; I’ll edit accordingly.

In my comments on the 2008 Gartner Data Warehouse Database Management Systems Magic Quadrant, I observed that Gartner’s “completeness of vision” scores were generally pretty reasonable, but their “ability to execute” rankings were somewhat bizarre; the same remains true this year. For example, Gartner ranks Ingres higher by that metric than Vertica, Aster Data, ParAccel, or Infobright. Yet each of those companies is growing nicely and delivering products that meet serious cutting-edge analytic DBMS needs, neither of which has been true of Ingres since about 1987. 

The general list of “market forces, end-user expectations and vendors’ resulting solution approaches” at the top of the 2010 Gartner Data Warehouse Database Management System Magic Quadrant article is a mixed bag. Following Gartner’s order, I’ll address those first, and particular companies cited afterwards. Specific items and comments include:

Long though it is, that list of general items and issues for the 2010 Gartner Data Warehouse Database Management System Magic Quadrant has some gaps. Most glaringly, I don’t see any references to advanced analytics in general, or even to the specific case of integrated predictive analytics. There’s also nothing about solid-state memory or other storage-technology considerations, although in fairness it’s still early days for much of what vendors conceive of as competitive differentiation in those respects.

Here are some vendor-specific comments on the 2010 Gartner Data Warehouse Database Management System Magic Quadrant:

Comments

23 Responses to “Comments on the Gartner 2010/2011 Data Warehouse Database Management Systems Magic Quadrant”

  1. zman on February 5th, 2011 11:31 am

    Gartner describing infobright as the “only open-source column-store DBMS on the market” is not just an exaggeration it is flat out wrong. Calpont/infinidb and DynamoBI/LucidDB are two others just off the top of my head. It makes me wonder how thorough Gartner’s research really is.

  2. Curt Monash on February 5th, 2011 11:47 am

    Actually, I’d have started with VectorWise, which they actually wrote about.

  3. George Chen on February 5th, 2011 1:39 pm

    Curt, very good coverage.
    Should be included by Gartner on same page or pdf where their MQ is. Hope Gartner folks will consider it seriously if they do care about being professional analysts 🙂

  4. Eric Kraemer on February 5th, 2011 8:22 pm

    Hello Curt,

    I really don’t follow the logic of your unflagging support for on-site POC’s. Consider that,

    1) A preference one way or the other most certainly favors certain vendors. Something that, from a customer perspective, is problematic from the start.
    2) If you don’t trust the vendor to run an honest POC, you shouldn’t be evaluting them to begin wtih.
    3) If you don’t have the technical know-how to ensure an honest evaluation, you should be hiring someone who does to the run the POC.
    4) It’s time spent that could be better used running a good POC – If its not time spent then see point #1 about favoring certain vendors.

    I’ve seen very good and very bad POC’s run both onsite and off. I challenge you to take a position on this topic as a customer, rather then vendor, advocate.

    Regards

  5. Curt Monash on February 5th, 2011 9:00 pm

    Eric,

    Your comment is almost entirely nonsensical, with the last sentence being highly insulting to me.

    If one only buys from vendors that one is certain would never engage in any kind of misleading sales practice, it’s pretty hard to run an IT shop beyond the “abacus” generation of technology.

  6. unholyguy on February 6th, 2011 12:34 am

    @Erik I back Curt up on this, as a customer. It is always preferable to run a PoC onsite if possible

    1: Your data is safer
    2: At an offsite PoC you have no idea how much effort went into tuning and optimizing the environment. This is important information as YOU are the one who is going to stuck doing all that at some point
    3: You get to see how the hardware performs in your data center, on your network, which is after all, where it is going to end up eventually
    4: Your team builds a lot of hands on, working knowledge about the product during the PoC which can help you make the final decision
    5: Vendors do, in fact, upon occasion, lie their asses off in PoC’s

  7. Columnar compression vs. column storage | DBMS 2 : DataBase Management System Services on February 6th, 2011 4:23 am

    […] getting the increasing impression that certain industry observers, such as Gartner, are really confused about columnar technology. (I further suspect that certain vendors are […]

  8. Curt Monash on February 6th, 2011 6:54 am

    I should add that it’s not ALWAYS the case that one should do an on-site POC. If there were general industry knowledge that the results of Oracle-controlled POCs were reliable guides to production, or at least to the results of customer-site POCs, then prospective buyers might well individually choose to not bother insisting on an on-site test.

    But as matters stand, Oracle is way too controlling of the Exadata-related information flow.

  9. Gartner’s 2008 data warehouse database management system Magic Quadrant is out | DBMS 2 : DataBase Management System Services on February 9th, 2011 4:45 am

    […] 2011 edit: I’ve now commented on Gartner’s 2010 Data Warehouse Database Management System Magic Quadrant as […]

  10. Gartner 2009/2010 Data Warehouse Magic Quadrant comments | DBMS 2 : DataBase Management System Services on February 9th, 2011 4:47 am

    […] 2011 edit: I’ve now commented on Gartner’s 2010 Data Warehouse Database Management System Magic Quadrant as […]

  11. Vendor segmentation for data warehouse DBMS | DBMS 2 : DataBase Management System Services on February 9th, 2011 4:48 am

    […] 2011 edit: I’ve now commented on Gartner’s 2010 Data Warehouse Database Management System Magic Quadrant as […]

  12. Gartner Business Intelligence Platform Magic Quadrant « In(tegrate) the Clouds on February 10th, 2011 12:42 pm

    […] Comments on the Gartner 2010/2011 Data Warehouse Database Management Systems Magic Quadrant (dbms2.com) […]

  13. DbGuy on February 10th, 2011 4:16 pm

    What exactly do you mean by “non-query analytics”? Google returns total of 4 results for this expression – all to this page or aggregators.

  14. Curt Monash on February 10th, 2011 4:55 pm

    Examples of non-query analytics: Data mining/predictive analytics, monte carlo analysis, etc. I.e., tasks where the main processing load isn’t just query execution.

  15. Eric Kraemer on February 11th, 2011 12:41 pm

    So my post is “nonsensical” – rebuttal in one word is clever indeed – yet you jump to the complete other end of the spectrum with this:

    “If one only buys from vendors that one is certain would never engage in any kind of misleading sales practice, it’s pretty hard to run an IT shop beyond the “abacus” generation of technology”

    So 1) All vendors are devious and 2) All customers are clever in their own house yet somehow take leave of their senses when they step on another vendors premises? – Ok that was a bit tongue-in-check – enough arguing the extremes.

    I’ve been a part of, or run, well over 100 POC’s of meaningful size over the past 10 years. On average the ones run in customer sites are the least well managed and the most prone to manipulation. Customers who run POC’s onsite never show up – they are dragged away by constant interruptions, delegate important tasks away and frequently have to deal with multiple vendors “cooks in the kitchen” onsite at the same time to make the logistics work. On a whole, its a mess. If you want opportunities to manipulate a POC – host it at the customore site.

    The very best POC’s I have seen have brought the customer teams to an offsite location and been heads-down on the technology for 3-5 days non-stop. If you are truly worried about vendors doing sneaky things (which is healthy but overplayed these days) bringing them onsite is a false promise of security. I’ve caught other vendors posting results that defy physics or prove optimizations were used that were off-limits both onsite and off – The customer is nearly always oblivious.

    The real defense is to have a well constructed POC with smart individuals participating. Onsite or offsite is a very small factor. The only time it should matter is if security reasons demand it.

    I still think your perspective here is biased. It favors the ability to efficiently deploy onsite POC’s and I can think of no good reason how that correlates to a “good DW technology”. One vendor in particular has made their entire POC strategy “Fear-mongering over onsite POC’s” because they have a deployment advantage that really on matters in POC’s (nobody deploys the hardware every day).

    It seems you have taken this to mean that I infer you were paid to take this position. I am not. These thoughts are exclusive, yet I could have made my position more clear. You have my apologies for that.

    I do think you over-emphasize onn of the least valuable elements of a POC – where it is at. And I think the points I have made supporting that are far from “non-sensical”.

    Regards

  16. Curt Monash on February 11th, 2011 1:44 pm

    Eric,

    I don’t respond well to comments that more or less accuse me of corruption. And while we’re speculating about people’s biases, honest or otherwise, it would have been nice if you’d more clearly identified yourself as being from Microsoft (you happened to have left a typo in your email address the first time around, so at that point I wasn’t sure that you were).

    Anyhow …

    I’m not a believer that all POCs should be done onsite, the way Netezza wishes that I and everybody else would be. But I believe that a vendor should at least have the capability of doing onsite POCs. Microsoft, I gather from your comment, does them. Vertica and ParAccel do; indeed, one can just download Vertica and do one’s own, and the same is true for Greenplum, Infobright, and others. Teradata does, despite having all the same apparent reasons Oracle does for disliking them. Netezza of course does. Oracle stands pretty much alone in doing onsite POCs so very, very, very rarely that nobody can identify an example of one.

  17. Eric Kraemer on February 11th, 2011 6:45 pm

    In fairness I should have disclosed that I still have a prsentation of yours in mind from…gosh probably 2 years back now..where you really hit the onsite POC topic hard (my perception). I think that one was TDWI related….

    And I am a bit off-topic in this thread – Your coverage of the Gartner “analysis” is about vendor feature/capabilities and the topical issue is who supports what types of POCs – as you pointed out.

    I should have dug up that old post!

  18. Curt Monash on February 12th, 2011 3:54 am
  19. Mike Pilcher on February 13th, 2011 4:17 pm

    Curt thank you for highlighting some of the things in the Gartner report. Over recent years a number of analysts decided SAND was going to be bought by SAP and stated this as a high likelihood. This has never been a SAND’s plan and so SAP buying Sybase didn’t faze us at all. It does seem to have caused analysts to question our strategy. This is a challenge as we are put in the position of trying to prove a negative.

    There are a number of other things worth addressing in the report. Highlighting SAND as a niche vendor when compared to Oracle, Teradata and IBM is fair. These vendors all have billions of dollars to be jack of all trades in a broad market. SAND’s focus is to be the best analytic database possible, and to focus on powering applications that require what you describe as investigative analytics, what are also identified as Ad Hoc, or Data Mining applications. This our focus. We do have customers with hundreds of terabytes and tens of thousands of users on the same application, but that is not our primary focus. Our focus is powering applications that require iterative analysis and to do that better than anyone else; churn analytics, billing analysis, marketing segmentation, retail market basket analysis. SAND is the database power behind the programme that many commentators say took one customer from the UK’s number 2 grocer to the number 1 in Europe.

    There were also a number of things we will ask the Gartner team to address in next year’s MQ because we must have failed to communicate some data points correctly, as follows:-

    “SAND has been in existence for approximately eight years” – we have been in business for 25 years.
    “Its technology is used as an analytic engine and as an archive engine” – we have customers who use us to mix workload exactly as Gartner describes. We have a column oriented database for ultra fast analytics and we have a column oriented extension for keeping vast amounts of low-value data online. We have customers who use us as an EDW, and as a performance layer for Oracle, SAS and others.
    “It will continue to struggle against the larger vendors and venture-funded start-ups that can invest more in R&D, marketing and sales” – we have 800 man years of development, I think the start-ups will have to run fast to catch up.
    “SAP’s acquisition of Sybase poses a technological challenge for SAND” – agreed with your comments, this is not a technological challenge for us, nor a business one. SAP’s acquiring Sybase at some level validates columnar technology, SAND’s in-memory strategy which HANA is very similar to, and it poses no more of a challenge for us then the “venture-funded start-ups”.

    There are some things we are delighted they highlighted
    “Because of the tokenization and column store, it requires no indexing or query tuning” – this is detailed as a key criteria for the future of warehousing, with Gartner acknowledging that others don’t have it. SAND has been perfecting this for over a decade (please see above the 8 year comment). So ironically this was seen as visionary then but we didn’t tell them, so that was our fault.
    “It is also a good choice for analytic data marts to support the off-loading of workloads from an enterprise data warehouse. In addition, several customers use SAND’s technology as an enterprise data warehouse” – Agreed, please note these customers that use SAND as an EDW are running hundreds of Terabytes of data and unusually in the EDW market deployed to tens of thousands of users, with use of SAND in real-time call centres with real-time responses for thousands of concurrent users.

  20. Neil Raden on March 22nd, 2011 9:46 am

    Curt,

    I’m engaged in an on-site Oracle Exadata POC for data warehousing. This is obviously the exception, but apparently it does happen.

  21. Rune Henriksen on June 1st, 2011 3:49 am

    Mike Pilcher; To call Teradata a jack of all trades is unfair since they are the only mega vendor here solely focusing on analytical data warehouse solutions and do not have any OLTP solutions like Oracle, IBM. This is their strength since they don’t try to get a database to do things it hasn’t been designed for at all like the other vendors here.
    But you are right that they have the funding and organisation in order to deliver globally and are not likely to be aquired by any of the other vendors.

  22. Dedupe Software on July 9th, 2011 2:23 am

    I’m also engaging in an on-site Oracle Exadata POC for data warehousing and apparently it does happen.

  23. Comments on the analytic DBMS industry and Gartner’s Magic Quadrant for same : DBMS 2 : DataBase Management System Services on February 8th, 2012 12:17 pm

    […] for Data Warehouse Database Management Systems is out.* I shall now comment, just as I did on the 2010, 2009, 2008, 2007, and 2006 Gartner Data Warehouse Database Management System Magic Quadrants, to […]

Leave a Reply




Feed: DBMS (database management system), DW (data warehousing), BI (business intelligence), and analytics technology Subscribe to the Monash Research feed via RSS or email:

Login

Search our blogs and white papers

Monash Research blogs

User consulting

Building a short list? Refining your strategic plan? We can help.

Vendor advisory

We tell vendors what's happening -- and, more important, what they should do about it.

Monash Research highlights

Learn about white papers, webcasts, and blog highlights, by RSS or email.