Quality of Documentum Over the Years


I recently received an email from someone whom I will call…Socrates.  He asked a question and I wanted to share it for discussion publicly.  First the question, then my reasons for the public discourse.

Laurence, I have been working on Documentum since version 2. I am now working on DCM 6.5 sp3. I find that the quality of the product is going down every release. What do you think?

The reason that I am bringing it up publically is because I don’t have a clear-cut answer.  As with products from most vendors, some releases are better than others.  I also only have direct experience with Documentum since the 4i release at the end of ’99.  To top it off, I haven’t used every component, much less every component of every release.

Of course, I have some concerns.  I saw Rick Devenuti speak at EMC World and he seemed preoccupied with addressing quality issues.  Whether these are long-standing or new is something we can discuss at the end, where I have a couple more thoughts.

In between, I am going to share some of my “quality” stories here, both good and bad.  I’m hoping that Johnny, Scott, Lee, and Robin all chime into the conversation.  Please do so yourself.

Remember, there is no “right” answer.  We are merely looking for experiences.

Life with 4i

imageI entered the Documentum world less than a month after the release of Documentum 4i.  The “i” should tell you all you need to know about the timing of the release.  I am hard pressed to describe the highlights of the release, but there are a few a gleaned while working with my colleagues who were old hands at EDMS98.

  • Workflow: It was new in 4i, replacing the old router method.  It had some issues with larger, more complex, workflows, but it was also the 1.0 version of a major feature.  Documentum worked hard to get it fixed, but I remember old Bob cussing at the machine when it would blow-up.
  • RightSite: Was better than EDMS98, but man did it have limits.  This wasn’t a quality thing though, just a limit to the technology and design.  All web interfaces were pretty primitive back then.
  • Goodbye WorkSpace: That desktop client was an old standby.  The install was kept around by Documentum techies for years and used until the old DMCL library was removed.  That shows a lot of quality in WorkSpace and in the backward compatibility of the DMCL over the years.

That is my baseline.  Interesting days.  The Workflow issues made me worry about quality, but back then I was more concerned with learning the complexity than dealing with the quality.

Carving a Path to 5.3

There were some basic iterations of 4, but with the 5.x product, there were some issues.  I didn’t deal with a lot of them as I waited until 5.2.5 to put it into a real production environment, but forget 5.1 and 5.2.  There were a lot of general issues.

There was a lot going on in this release.  Everyone’s favorite was the new Web Development Kit (WDK) and the growing usage of the DFC.  I think the Java Method Server may have been new in the 5.x release, but that is a little fuzzy.  If anyone knows for sure, please share.

5.2.5 was okay, but 5.3 was a total nightmare.  image Forget the core product, the issue was the new Index Server.  FAST was “fast”, except in getting it to work correctly.  There was a large difference in the wilds of the data center from the clean world of the Documentum test-beds.  It took several service packs to get it right.  I think SP3 was the SP where you actually had to blow away your index and start over.  The lessons learned from this debacle have led to a much more conservative course for releasing the new Enterprise Search Server.  The slow pace to release is frustrating, but so was search blowing-up in production.

By 5.3 SP4/5, life settled down.  Since then I don’t think I’ve upgraded because I had to upgrade, only because I wanted to go ahead and do it.

Which brings us to the world of 6+…

Attack of the D-Versions

Starting with a large number of presentations in 2007 talking about D6, every version has been referred to as Dx.x.  I think some people in the marketing department wish they hadn’t let that one hit the slides at EMC World 2007.

Aside from that, I’ve been following a simple approach, only upgrade to SP1 or higher of any version.  Since I’ve done that, I’ve only had two real problems.

  • LDAP Synch: To be fair, this is suffering from old age.  They have spent a lot of time trying to fix it, but I keep having to find all sorts of new ways to work around it.  It works great for smaller user populations, but when you start to cruise past the 5,000 mark, things start to become fun.
  • Federations: This isn’t a loss in quality. This stems directly from the fact that the Federation process hasn’t changed in 10+ years.

Now, I know that there have been problems here and there.  I know the Branch Office Caching Server had some issues when it first came out.  I also know that most of the products that I see having issues are usually shinyimage new “1.0” products.  The core Content Server has been doing fine, as have many other products that are just “evolving”.  While it is a shame that you don’t want generally want to install the first release of a new product, that has actually been consistent for years.  I also use the same approach with Microsoft and other major vendors as well.

There is a lot to test, and a lot of permutations in the real world.  There will always be things that aren’t found in testing because you and I will always be throwing these products into unclean, old, cluttered repositories that EMC just doesn’t have lying around.

So the real question is two parts:

  1. Have you seen lots of issues in existing products that seem to be creeping up in each release?
  2. With new products/major features, have they been more problematic or do they have the same (or less) issues than previously released products.

Other food for thought…was Rick harping on fixing new quality or old quality issues?  I suspect old.  Is the “focus” on quality just typical marketing, realization that they need to fix it, or something they are going to fix instead of innovating further?

Let’s figure this out…

16 thoughts on “Quality of Documentum Over the Years

  1. Interesting discussion point. My background is a little bit from 4i but mainly from 5.x, and even then has been in and out. I agree with what you say about the problems tending to be with the ‘newer’ products, this is backed up by some recent experience with the SharePoint integration products (however if you expect there will be some issues your much better prepared to face them).

    The reason why quality may be declining is the sheer breadth of products which are in the suite now, just consider how many different client products there are! Some of these are bought in products and then rebadged (e.g. TruArc -> Records Manager) and some are new products developed in house (Centerstage). Speaking of Centerstage this product has been delayed, the reason quoting is getting it right, this is reassuring to hear and becomes a case of getting the messages out at the right time for the release (there has been way too long a lag with Centerstage).

    Overall though I would say the quality of the core Content Server has moved on, considering the amount of functionality which can be exploited through the various components. Just be careful how you plan for projects with new, or even relatively new, products and make sure you put the support channels in place before you need them.

    Like

  2. From within EMC I can see that there is huge push towards quality of the Documentum product suite – it’s not just marketing hype. The product groups are now directly accountable for quality and this will have a huge impact on the future releases.

    In defence of product quality, many of the issues are due to the sheer range of functionality and the myriad of platforms and components that it now needs to support – It is impossible to ensure that all permutations are tested, much as we’d like them to be.

    In Australia we have seen the number of product and project issues declining as we become more prescriptive around the recommended configurations – we’re a consulting organisation and we’re expected to offer best practice advice, in the past maybe EMC has not been firm enough around our recommendations….

    EMC is taking product quality seriously, I hope this starts to be reflected in future releases.

    Like

    • So that is a vote on improving the quality going forward. As I stated, I understand the challenges involved with all the various options.

      The question still is, is the focus on quality because of recent issues or because of ongoing concerns? [not a good way to say this without sounding critical].

      Like

      • The focus on quality I believe is historical more than recent concerns and some of it is aimed at addressing the perception of quality.

        I’ll share with you my experiences of when I joined EMC (from Oracle)

        There was a perception of poor products and implementations leading to dissatisfied customers, poor references etc. etc. etc.

        I dug into each and every one of these and although there were some product issues a lot of it was related to CMA Consulting (or the partner) not being prescriptive enough about certain things.

        Some of it was recommended configurations not being adhered to, some of it was a lack of business requirements, compounded by a lack of an information strategy, project management was seen as an overhead and the amount of post -production support required was woefully under-estimated and some of it was just the drive-by shooting approach of sales…

        In ANZ we’ve fixed all of the issues within our control and 80% of the issues have gone away.

        All I can say is that I see and sense a real drive toward quality under Rick and hopefully this coupled with the focus on Consulting (we don’t use PS anymore) from David Gai will allow us to shake the flaky image to the benefit of our customers and the whole Documentum eco-system.

        Lawrence

        Like

  3. I’ve been working with Documentum since 5.2.5 and generally share the same opinions on the releases. I have my list of significant troubles starting with XML applications to LDAP and object replication but I would want to underline the matter of “perceived quality”.
    The user experience keeps getting down and that’s a major concern since it undermines one of the pillars of ECM adoption: if the users are not comfortable then you have a challenge to build more on the platform.
    Currently I’m building and implementing custom clients. They perform better and address the user needs better. My suggestion to EMC would be to put a lot of commitment in building a better Enterprise user experience – this will raise the perceived quality and help a lot.

    Like

  4. As another veteran from EDMS98 days, I concur with what piewords summarizes. Luckily there are (unofficial) “safe harbor” versions that customers can land on in between the various other releases. The problem with those is that by the time they get established (e.g. 5.3 SP5+) the End Of Life for the version is drawing near. It might be a strategy for EMC to consider to allow those to continue for longer. One specific example: Composer/DAR Installer for 6.5 have known problems, and (before 6.6 which I havent tried yet) very poor error logging when installing DARs. The alternative was the stable DocApp Installer 5.3 SP6.5 to install DocApps – but before the DAR installer was really fixed and made useable, the DocApp Installer was end-of-lifed.

    Like

  5. Allen says:

    Starting worked with Documentum from 4i days, I just want to share my experience regarding the ‘quality discussion’. I finished both major upgrade from 5.2.5SP3 to 5.3SP3 and then to 6.5SP1. The experience to 6.5SP1 was much, much worse than the first one. You would see many issues here and there. Part of the reason was from the new features/components introduced, and most part was due to lack of rigorous quality control (you could see windows path used with UNIX installation).

    Besides the qulity issue within the product suite , I have also seen the lack of troubleshooting knowledge and skills from the support team (is it part of the product quality? I think so). An extrem experience for me was a service request regarding indexing. The case was closed after 9 MONTHS even without my explicit consent. You would guess right. I never got the issue resolved.

    I agree with James’ comment on Composer/DAR. It’s never a good thing if you provide half-cooked meal. The hype of Composer to replace DAB (was it fully materialized?) , and then the delayed Centerstage release showed off-balance among product definition, quality control, and delivery.

    Like

    • Allen, thanks. One thing about Support, it has been a constant over the years, and to be fair, I’ve had similar experiences with other Content Management vendors and their support line. I think part of the problem is that when “experts” break down and open a case, 50% of the time it is well beyond the capabilities of the Support team.

      Like

  6. Chris Campbell says:

    I’ll try to make this first comment brief. (Still catching up from vacation, injury, work, etc.) The question from Socrates really is too open-ended to have a proper discussion. I’d like to know in which areas does he see the quality dropping? Integration? Usability? Bugs? Incomplete features? It’s difficult to fight against or for such a shadow.

    Since I can’t address Socrates, I can only rely on personal experience on what I perceive as change in product quality over the years. Things have changed since “The Good Ol’ Days” for both the better and the worse. Overall, I’d have to say that quality has *generally* improved over the years. In my final estimation I take into consideration the growth of the product and the company; and also what I consider the fringe elements like the user community, user support and sales staff.

    I’ll quickly touch on the actual questions to be answered. 1) Current issues that seem to be creeping up in each release? Simply, yes. There are *still* core issues in the ORM and the DFC that I’ve identified and passed on to product managers. Architectural items like that aren’t quickly fixed and impact the *entire* product line. I get that. One particular issue will probably require significant resources to fix and a EMC VP to champion. Maybe I got lucky and over the summer it was fixed in 6.6.

    2) New releases – Better or worse than in the past? Here’s where I point to where quality has improved over the past. From my usage and viewpoint new features work as promised and don’t cause other items to break. This is a *major* improvement over the 4i days when I had to keep a matrix of which patches worked with each other and which ones broke others. Quality can always be better (Praise Be Unto Deming). I see EMC as not just treading water, but slowly starting to move forward.

    This would be a good moderator topic for a live discussion so we could direct our focus onto specific areas that people feel need higher quality.

    Like

    • Thanks for your “brief” comment Chris. 😉 What we have so far is a mix of opinions. So far we agree that there has been no “ideal” release in the last decade, though some SPs rock. There is a belief that the recent focus on improving quality it real, but hard to measure at this point.

      This is an interesting topic if for no other reasons than it is colored by our personal perceptions over time. Looking forward to more people chiming in.

      Like

  7. John says:

    I have examples of areas in which I feel the quality has radically degraded or simply been ignored, potentially written off with bad excuses.

    Desktop Client -> Webtop: feature parity still does not exist. When I talk with clients who used DTC a lot, they still bemoan the absence of simple features like saving search results and a really easy integration into Windows Explorer OOTB. Sure, there is an {insert several thousand dollars} product that can be glued-on that does something, but it is still not DTC.

    Documentum Application Builder -> Composer: feature parity is far from there. Modifying an application once installed still requires use of DAB which has the exact same quirks I’ve seen in it since 5.2. Fix Release in PowerLink is blank for these issues.

    Documentum Application Installer -> DAR Installer: neither one is good about reporting or handling errors. These are THE ONLY way to install an application into another system. As the only installation mechanism, these two entities should work substantially better than they do.

    Workflow Manager -> BPM: Developers/Users either use technology from 1999 or pay $75K to get something that works properly. The way in which WfM has been neglected is awful.

    5.3 -> D6: The resource requirement jump is mind boggling. I remember running a development environment in a VM with 512MB of RAM. Now 1G, preferrably 1.5G, is needed. The bloat is unbelievable.

    ADTS: The silent failure during the AdLib install is a real hang-up for those not familiar with the process. Additionally, the outdated version of AdLib being used is hard to troubleshoot since it is (a) hard to find an EMC support person who knows about it and (b) AdLib themselves go into shock when contacted and told the version number. Excellent plan to use 3rd party tools (reinventing the wheel == bad), but bad plan in persisting with something very out of date and basically unsupported.

    Like

    • Thanks for your input. Good to see passion. I definitely agree that feature equality from old product to new product has been a major issue. I had blocked Composer from my mind because I didn’t use it for a while as I live in Application Builder instead.

      The bloat from 5.3-D6 is bad. Better with 6.5 when they switched to jboss. Would still like to know more about the reasons. As for AdLib, I’ve been lucky and not had major problems.

      No matter how you slice it though, this is definitely a vote towards the negative.

      Like

      • John says:

        Indeed a vote towards the negative. I avoided leading with that and listed personal experience instead as a more practical statement of my opinion.

        That being said, there are many neat things being done. Some I feel are a little late in the making (i.e., the D6.5 webtop interface), but progress is still progress.

        But, I have to track back towards the negative. Because I feel that unless there is improvement in quality, especially in the OOTB basic features, the recipe is in the making for Documentum to make itself obsolete in favor of much easier and readier things to implement. SharePoint, for example, is not a perfect replacement, but it is there and much more accessible to the end-users in many cases. Documentum’s quality needs to improve and such measures of end-user ease-of-use need to be implemented to keep Documentum viable. From a logical point of view, it is highly viable (I cannot think of anything else I would manage enterprise documents in, especially where workflow or records management is concerned), but it is the end-user perception that ends up winning out many times. Especially when that “end-user” has the word “Executive” in their job title.

        Like

  8. I started with version 2, back when I was just a newly-minted UNIX geek. One thing you missed with the transition to 4i was the introduction of the DFC. DMCL had a very UNIX feel; a simple, open API designed to be glued into any programming language. DFC was just Java then, with a COM layer growing over it later. That was also the point where EMC became more marketing-driven and started chasing the Internet bubble at the expense of their existing clients.

    Both were attempts to capitalize on hot topics of the time, Java and the Web. I never bought that the DFC would make a whole pool of talent available; Documentum’s about the model, not the means. However, the marketroids successfully reframed it. Hiring managers now believe they can take Java people and mold them into Documentum people, and I hear gasps of disbelief when I say Java or Visual Studio aren’t requirements to do Documentum–a good Java programmer is not necessarily a good Documentum developer.

    The lack of focus then is what brings us to the lack of quality now. Innovation at the model and server level is rare, and frankly I don’t give Documentum much geek cred anymore because of it. Great ideas like BOF and Aspects are stapled into an API rather than made an inherent part of the product. Too much work up the stack (and on vertical solutions) has made the product top-heavy and tottery. EMC continues to chase markets (i.e., case management) rather than concentrate on making a solid core product.

    Like

  9. BonScott1980 says:

    I have been working with DCTM since 2002, and have tried very hard to be a loyal supporter. Like many others, I have suffered through a myriad of quality issued but we always soldiered on the promise of better things to come. As has been mentioned before, each upgrade became more difficult and more costly. The most recent to 6.5 has topped them all. I’m not sure the return is worth the investment. I am quickly becoming an SP 2010 fan. I have been searching for reasoning why I should continue to be a DCTM bigot, but I can’t seem to find any with the advent of SP 2010. Am I missing something? Can someone tell me what features are offered by the DCTM suite that I cannot get for a lot cheaper in SP 2010 (or cheaper yet, Open Text)??? Is tiered storage for companies that have 5oo M+ docs? Is it DAM? I need some hope!!!! Or is it help??? Thanks….

    Like

Comments are closed.