Science Interactive Group Knowledgebase

Cloud Bug Status Updates

Updated on

This article contains details and up-to-date information about active and relatively widespread bugs affecting the Cloud platform. This page will be updated as we investigate and correct issues.

This page is not intended to be a comprehensive list of all maintenance being done on the Cloud platform, or all known bugs, or even all in-progress bugs. This article is reserved for high-urgency or new, relatively widespread issues that warrant general internal communication. As such, the list of issues on this page will likely be brief.

In Progress

Expand or collapse content UX Improvement: Slow load times on new gradebook
Description Load times are too long on new gradebook.
Priority Medium
Bug type UX optimization opportunity
Who's Affected? All users -- does not prevent progress or use
What's the cause? The huge amount of data that has to be loaded
What should I tell my customers? Optimizing the new gradebook loading experience is an ongoing project that we are actively working on. We expect load times to decrease in stages over the next semester as we find and launch new optimization options.
Updates
Expand or collapse content Data Tables showing as incorrect in the gradebook
Product Management Squad | Microsoft Teams
Description Answers in some data tables are being autograded as incorrect
Priority Medium
Bug type Visual
Who's affected? Extent unknown; appearing sporadically 
What's the cause? Learnosity's tools are setting some open-ended questions and data tables as autograded and marking the table answers as incorrect.
What should I tell my customers? We are researching the issue and working on a universal fix to remove the autograding alert from unintended questions. In the meantime, we can fix individual lessons right away if supplied with the course URL, lesson, and affected question(s).
Updates

Recently Resolved

Expand or collapse content RESOLVED: Student lesson lists are blank
Description Lesson lists for some students are blank
Priority High
Bug type Display
Who's Affected? Students in a selection of biology courses
What's the cause? Unknown
What should I tell my customers? We've discovered that, for some students in some courses, lesson lists have appeared blank. Student progress is not affected, as graded lessons are still present in the course gradebook. This means the lessons themselves have not disappeared, but the list page used to navigate to lessons is not populating. Researching and resolving this issue is the highest priority for our support team.
Updates (4/18/23 9:20 am MDT) The issue appears to be restricted to courses that contain the SI version of the lesson "Biological Macromolecules"

(4/18/23 9:00 am MDT) We have determined that the bug first occurred between 10:00 am and 1:30 pm on 4/17/22. These are the SKUs we have found to be affected:

SC-10001-BK-01
SI-10325-BK-01
SI-10478-BK-01
SC-10001-BK-01
 SI-10127-BK-01

Expand or collapse content RESOLVED: "Submit" button not working for some students
Description For some students, pressing "Submit" at the end of their lesson doesn't actually submit the lesson, it just loads the page again. When impersonating the student, we are able to submit the lesson with no issues.

First Report Date Friday, 2/10/23
Priority High
Bug type Browser update
Who's affected? Students using Google Chrome whose browser version updated to the newest version Google released on Friday.
What's the cause? The issue appears to have been caused by an update Google Chrome made on Friday, which caused errors in Cloud's ability to track and record student's progress through the lesson.
What should I do when I get a report of this? You can ask the student to try the submission in a different browser, or you can impersonate the student and submit the lesson for them. NOTE: this seems to only work using Firefox or Edge browsers -- it does not work using Chrome.

If you are unable to submit it via impersonation, submit a ticket.
Updates 2/14/23 8:12 am: The fix has been deployed live and we are currently validating with some affected students to be sure they are now able to submit.

2/14/23 8:05 am: We have a fix for the issue and it is currently being tested in staging environments. Assuming testing continues to go well, we expect to deploy it to the live site this morning.

2/13/23 2:52 pm: We have now learned that the issue is significantly more widespread than we knew, with over 200 tickets in Customer Service between Friday and today. We are preparing a banner that will go up on the platform as soon as our developers are back at work (around 2::00-3:00 am tonight), and they will be working hard to figure out what part of the new Chrome update caused the problem and building a fix for it ASAP.

2/13/23 11:18 AM: We are currently trying to fully determine the root cause of the issue and trying to see if there is a safe workaround we can build out for this busy semester. The long term fix for this and other red-circle type issues will require fully refactoring the student progress tracking logic, which is too risky to do during Spring or Fall semesters.
Expand or collapse content RESOLVED: UX Improvement: Autosaved Comment UX Issue
RESOLVED Fix pushed to Prod 2/2/23
Description
When typing a comment in the new gradebook, the auto-save feature prevents the user from continuing to type, causing annoyance when an instructor pauses too long while typing
Priority Med-High
Bug Type
Not a bug -- UX issue
Who's affected? Instructors using the new gradebook
What's the cause?
The way autosaving was set up
What should I tell my customers? This issue has been resolved and a fix pushed to production on 2/2/23
Expand or collapse content RESOLVED: Users receiving Error 5000 via LTI
First Report Date 1/16/2023 9:00am
Priority UPDATE: Low
Who's Affected? Some users at certain times -- the issue is neither universal nor consistent. When users try again after waiting a while, the issue nearly always is resolved on its own. We have seen significantly fewer cases of this instance in the past two weeks.
What's the Cause? We have determined that this issue is caused by the customer's LMS, not by SI Cloud, which means we are rather limited in our options for fixing it. The error occurs when the LMS sends an incorrect context_ID value along with the link-click, which means that Cloud doesn't know which course instance the user is supposed to be associated with and therefore doesn't know how to take them into a lesson or course.
What's the Timeline? Fixing this is in the hands of the LMS'
What Should I Do? If customers report this issue to you, gather the following information, then ask them to try the link again after waiting some time. If the issue persists, submit the info via ticket:
  • Section info (institution section URL, instructor, SKU, etc.)
  • Lesson name
  • Names of students affected 
  • Names of students who have successfully accessed the lesson without the error (if any)
What Should I Tell My Customers? So far, the issue always resolves itself if the user waits a while and tries again. We are working to identify the exact amount of time they should wait, but it appears to vary. We are actively researching a permanent fix and updating our database, but want to be extremely careful not to make any updates that might affect any other part of the platform, especially at this early point in the semester, so testing takes a while.

If students have urgent need of access, you (sales rep) can turn off the section's LTI restriction temporarily and have the students sign directly into their course using the Cloud sign-in page (myhol.holscience.com). Their username and password will be the same as their LMS credentials.
Updates (2/2/23 9:40 am)The issue seems to be resolved or nearly resolved on the LMS end -- we have received very few reports of this issue over the past several weeks. The issue will remain open, but has been moved to low priority.

(1/20/23 2:37 pm) We have determined that this issue is caused by the customer's LMS, not by SI Cloud. The error occurs when the LMS sends an incorrect context_ID value along with the link-click, which means that Cloud doesn't know which course instance the user is supposed to be  associated with and therefore doesn't know how to take them into a lesson or course. Unfortunately, since the error is caused by the LMS, it is much harder for us to do anything about it, but we are trying to figure out what workarounds we can use to get the user to the right place without that important Context_ID value.

(1/20/23 9:10 am) We have now had one instructor experience the issue. We continue to work on database improvements to try and prove our hypothesis on the root cause, but since we cannot reproduce the problem, making changes is particularly risky, so we are taking extra steps to ensure cautious development and thorough QA testing before anything gets pushed to the live server.

(1/19/23 9:30 am) The sporadic nature of this issue makes us believe it's related to network traffic over LTI specifically. We are updating the error page to instruct users to try again later and also investigating ways we can better balance LTI traffic.

(1/18/23 5:55 pm) We continue to receive sporadic reports of this issue, and it almost always resolves itself the next time the student attempts it, which makes it extremely hard to troubleshoot. We are still unable to reproduce it on our end, but the additional logs and records of students attempting access again both successfully and unsuccessfully are helpful in troubleshooting and this remains our top priority.

(1/18/23 7:30 am) We have received additional reports of this error occurring as of last night. The developers had already added additional technical logging for LTI connections earlier this morning and are currently combing through them to see if any patterns emerge. We are still unable to reproduce the error,  as it does seem to be totally random.

(1/17/2023 9:30am) All reports of the issue appear to be resolved. Continued reports appear to be leftovers from the holiday weekend and those students are no longer experiencing the issue. We are considering the active issue tentatively resolved, and will continue to investigate the root cause

(1/16/2023 3:20pm) We have received multiple reports of the issue resolving itself in multiple sections and for multiple students. This leads us to believe the cause of the issue could have been temporary network or domain disruption. We will continue to monitor the situation and work with customers to ensure that full functionality is restored to all users.

(1/16/2023 1:30pm) Progress is slow because we are having trouble getting details from customers, but we continue to try and reproduce the issue independently.

0 Comments

Add your comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Previous Article Why is a Student Already Unlocked?
Next Article How do I find the instance URL?