When it comes to software testing, whether it’s automated or manual, testers should look for the most appropriate solutions for each bug found.
Another month has passed, and we return with a new entry in our Bugs of the month series.
BUG #1: Double progress registered
Project context: An online platform that includes a learning management series. The platform contains a public profile page for each user, displaying: awards, points, certificates, and other similar data.
Issue: Our testers received reports of loading errors on the public profile pages. Through Frontend and Backend results, the latter showed a non-unique error for one of the objects it was querying. In our case, it was related to users receiving badges for completing a course. For users experiencing this issue, the QA team found a duplicated transcript which meant they were awarded the badge twice by the system.
The testers found a way to replicate this bug by using Backend directly and calling the API that saves the progress on a course. The reproduction steps were done simultaneously, multiple times using JMeter, with multiple Threads and a synchronized timer. They found out that due to a short timeframe (a couple of milliseconds apart between these API calls), the system would consider each call valid, thus creating duplicate entries.
Currently, the database is queried by testers, searching for duplicates that are being fixed manually.
BUG #2: “Shared search history”
Project context: An online platform with a search feature designed for clients who want to register to stay updated with products and services provided by the platform.
Issue: While testing the user login and logout option, we found that the search history is not deleted when the user logs out or when the session expires.
While the bug was initially dismissed as a minor issue, the QA team continued testing until they discovered that the history of the previous login session remained stored locally. Any user who logged on to the platform using the same device could access the search history of the previous login sessions. In short, it became a “shared history” between users accessing the same device.
After this comprehensive report was provided to the dev team, the bug was moved to an “urgent fix” category. The solution was to have any user search history deleted on logout.
BUG# 3: Files retrieved from outside the selected folder
Users can use different platforms for storing files online. While this proves advantageous for categorizing files and data, it can also lead to several issues, such as the one found by our testers.
Project context: A project which employs an online app used to retrieve company data from multiple sources, such as Outlook, Google Mail, SharePoint, Google Drive, etc. These files are checked for GDPR compliance documentation issues.
Issue: When checking for file locations, the testers discovered that the app began retrieving files from outside the selected folder. This issue occurred only for Google Drive. On closer inspection, the required files couldn’t be found anywhere in Google Drive.
After a couple of days, we figured out that the app was retrieving thrashed files instead of the files from the selected folder. To fix this issue, a filter was set to stop fetching the thrashed files.
Bug #4: Browser app connectivity issues
Project context: A desktop app and a browser app are intertwined.
Issue: If the network connection fails, the desktop app maintains some functionality after losing the network connection, while the browser app loses all its functionality. Suppose the user wants to navigate from the desktop app to the browser app during a period of no connectivity. In that case, the browser app will have its text replaced with code until the network connection is re-established.