SCORM vulnerability to networks

We have some courses created in Storyline that test well. However, we notice that some SCORM information is not making it to our Poeplefluent LMS about 1% of the time. With tens of thousands of users, this adds up to a lot of lost bookmarks and coure completions of "in progress" when learners have completed courses.

Our LMS is hosted so people are taking the courses from their homes and workplaces across many locations. I've noticed the same issue with SABA.

I know that SCORM doesn't always work well with Citrix and VPN and is vulnerable to things like firewalls, proxy servers etc, but what is the overall expected reliablity of SCORM for these intermittent, non replicable issues?

How vulnerable is SCORM to network traffic and what do I tell my stakeholders when they ask me to fix this issue. Whenever I go back to specific user, they can't replicate the issue so how do I resolve it?



8 Replies
Gerry Wasiluk

Hi, James!  Not sure I've ever heard of a "reliability number" for SCORM.

Too many variables IMVHO:

  • The LMS itself
  • The published content
  • The network
  • Intermediate servers
  • The user
  • The user's device
  • The browser and add-ins it's using, etc. etc.

The "dirty little secret" in the e-learning industry is that content always works with a LMS.  After integrations, content problems are usually the biggest problem with on-going LMS operations.

And all too often, I've seen LMS companies always blame the content.  That's too simplistic sometimes and often very wrong.

The problem with finding the cause for most content issues is that it is often like investigating a fire:  You have to deal with the aftermath of the remains of whatever you can find to get a possible picture of what may have happened.  You sometimes feel like you need monitoring equipment on the learner's PC to track things when the problem occurs.  Not likely.

It's nice when you can access the LMS DB and write queries directly on the data.  Sometimes the LMS stores more data than is available in any canned reports.  I used to write (read-only) SQL queries all the time against our LMS's Oracle DB.   Once I understood the LMS data model for content, and what was being stored that might help me out, I often found clues in that stored data that led to a likely idea of what happened for some issues.  But not all.

When we started using Articulate Presenter with our Saba LMS, we had a completion failure rate of over 20% outside of the corporate headquarters.  We were just using AICC then.  Turns out our Saba system couldn't keep up well with the consecutive posts from the content to the LMS at course exit  and blew posts off if it was still processing another post for the same learner and registration.

Articulate "rescued our bacon" and got us this solution and our failure rate went down to under .1 %--but we still had failures.  I think they are unavoidable.  You have LMS software from one company working with software produced by another company (like Articulate) that is running on hardware, software and other technologies from many other companies and being served on a plethora of learner devices and software platforms and browsers. Some days I'm surprised anything works.

You need to do your best to make things as best you can, investigate failures when you can and it makes sense to do so, and then accept that you can't solve everything. It's the old Serenity Prayer:"God, grant me the serenity to accept the things I cannot change,
The courage to change the things I can, And wisdom to know the difference."

I've seen so many weird things I can't remember them all.  Remember having a sales employee in Italy having issues.  For all the other folks in his office everything worked fine.  Only he had problems.  Everyone had a stock company PC supposedly only running the company-supported software.  When we asked him to try the course on a different PC, one where the course had formerly worked just fine for another person, and the employee's problem then "traveled" to that PC.  Weird . . .  almost like his credentials may have been an issue  . . .  We even put a trace and monitoring on as he took the course.  No clues.

Some of the more common things that I've seen:

  • Network latency or LMS load issues cause commands to execute in an unintended order yielding unpredictable or undesirable results
  • Network session timeouts
  • The LMS is being stressed and not properly sized to handle a heavy load
  • SCORM or AICC is not properly set up in the LMS
  • The company producing the LMS interpreting AICC or SCORM standards far "more strictly" than required or interpreting them differently
  • The learner's session with the LMS times out but the content keeps getting delivered
  • Something on the learner's PC prevents the course from working correctly
  • Pop-up blockers
  • The learner never really finished the course--they exited early thinking they had finished (managers and executives seem to be notorious for that)
  • Firewalls blocking communications
  • Content not working well in the LMS player frame for content
  • A “Keep alive” is issued too frequently (e.g., 30 seconds)putting a heavy burden on LMS which might result in unpredictable or undesirable latency-related results (Storyline's default keep-alive is every 10 minutes, though, at course exit, many posts happen in short order)

One more possibility: - if I was using AICC or SCORM 1.2 I'd keep my courses very short with shorter quizzes and not too many interactions.  Otherwise, large courses with many quiz questions and interactions could possibly cause tracking issues in the LMS.  Break large courses into a set of smaller courses or do multi-AUs or multi-SCOs.

Also, it could be possible there is a bug in the authoring tool and its output.  I've seen this a couple of times before, even in Articulate content.  To their credit, and from my experience, Articulate has always been great to address the issues once I could properly demonstrate and document them.  Couldn't have asked for a better response.

Good luck going forward.  LMS and e-learning content issues are always "fun"--as in aspirin-inducing and hair-pulling.  Hope something in this reply may help.

Steve Flowers

We've seen issues with sessions dropping but this is an LMS issue. For example, in the middle of a course I could pull the network cable for a minute to simulate an internet outage and the LMS would lose the session and tracking may or may not restore when the connection is restored. I think this si the cause of most of our failed tracking issues with our LMS. The content doesn't usually catch it when a connection is lost.

Victoria Aleman

Looking for something else on the Community, I stumbled upon this post thread and THANK YOU!!!  "Where have you been all my time and my previous co-worker's time dealing with course completion issues and "dropped communication" between the course and the LMS!"

Thank you, James for asking your question, and thank you, Gerry for your detailed answer.  We may finally get somewhere with addressing our issues.


Stephen Cone

There is one item I did not see listed in Gerry's comprehensive posting as a potential issue, namely:

  • Since SCORM supports the usage of JavaScript, you may enoucnter security problems with cross-domain scripts.  For example, if your content is located on and your lms is on, you'll need to ensure that both servers support cross-domain scripting to ensure that there is no breakdown in communication.

This is primarily an issue when dealing with third-party hosted content and proxy-servers etc.

Just something else to consider.

Philip Siwinski
James Andrews

With tens of thousands of users, this adds up to a lot of lost bookmarks and course completions of "in progress" when learners have completed courses.

Hello heroes,

I was wondering if any of you managed to solve issues related to lost bookmarks and course completion, specially with a large number of learners.

I'm experiencing such issues without being able to reproduce the issue. I suspect a mix of network, VPN, SSO might be responsible. Both Articulate and the SaaS LMS vendor are not able to reproduce the issue neither but it keeps happening "randomly" to many learners (which is a small percentage of all learners).

Any thoughts?

Thank you 😀