6.1 Support Solutions and Known Issues

Solutions and known issues to specific issues and scenarios that may arise from Zephyr 6.1 Release.

Zephyr Component	Issue/Scenario	Issue/Scenario Result	Solution
Server	After leaving and running the Zephyr server machine idle for more than 6+ hours (ex. 12 hours).	Server may receive crash throwing a "Too many open files" error.	The number of open files needs to be sufficient due to moving the transport node of elastic search as it requires more sockets for communication. Increase the ulimit on the server to 65535 to provide more sockets for communication. Setting ulimit in Linux: Stop the Zephyr Service. As the root user, open the file /etc/security/limits.conf for editing and add a line at the end (replacing <zephyr_user> with the user used to install Zephyr). <zephyr_user> soft nofile 65535 <zephyr_user> hard nofile 65535 Start a new session and then start the Zephyr service.
Server Upgrade	After upgrading from 6.0 to 6.2 in oracle11g.	Services are not started automatically.	The service startup is triggered by the upgrader but there is the response in success because the actual upgrade happens at the time of the server startup and it requires time to upgrade depending on the data. The services need to be started manually after performing the upgrade from 6.0 to 6.1.
Jira Sync Requirements	When a mapped Jira Sync requirement is moved to another release after being setup with previous JQL in both releases.	The requirement ID link in the "Testcase Coverage Details" window navigates to a blank page of requirement details when the mapped Jira sync requirements is moved to another release.	During the Jira Sync, while having overlapped requirement (a requirement falling in multiple releases) results for a matched JQL and having a different folder structure in both releases, the requirement is only displayed in the latest release sync. If the requirement existed in a previous release, it would be displayed only in the latest release sync, as the result of the this, test cases to requirements mapping from a previous release would still be shown in the UI. Whereas, the requirement is really moved to a different tree which no longer belongs to the source release (requirement internally mapped source release as well, hence the test case to requirement mapping is still active),
Jira Sync Requirements	When changing the Jira connection on a project after previous importing Jira requirements.	The previously imported Jira requirement link changes for the Jira URL.	Jira links will be broken once a connected Jira on a project is modified with a new Jira connection. The Jira links are made dynamically by finding the Jira url and issue key. Once it is connected and the Jira is modified, the link will be constructed with a new Jira url where as the key still belongs to the older Jira connection which will result in a wrong url. It is recommended that before a user switches a connected Jira from a project, they fully delete the imported Jira requirements from the previous Jira and clean up the data to avoid such behavior.
Jira Sync Requirements	When using a different JQL with the same folder structure across two projects and after updating a requirement from the first project to satisfy the JQL in the second project.	Updating the requirement from the first project to fit the JQL of the second project should display the requirement in the second project but the requirement will not display.	This can be prevented as we have a different JQL with the same folder structure in 2 different releases of the same project.
Exploratory	When first installing or reinstalling the Zephyr Exploratory chrome extension plugin on the web browser.	In manual mode of the exploratory plugin, users trying to use the full screen capture screenshot will take a screenshot but it will not capture the entire screen.	After first installing or reinstalling the Zephyr Exploratory chrome extension plugin, simply refresh the browser to be able utilize the full screen capture function if the problem persists.
Exploratory	When clicking the exploratory icon while in the edit window after clearing all the screenshots.	The exploratory plugin widget in chrome made display a cutoff interface and might not display the full interface.	If the user is running into this issue, simply click the exploratory plugin icon in the top right of the browser.
Vortex and Test Execution	After creating a Vortex job and executing a job with test case executions that have attachments/logs for duplicate files in the folder.	The attachments/logs are not attached to failed test cases in the test case execution when there are duplicate files in the folder.	If a user runs the same Vortex job multiple times then it works fine even if the result file is the same. To prevent the scenario from happening, users should not run the automation jobs having the same result file duplicated inside a folder. The following scenarios may work: If separate jobs are created for each file If separate jobs are configured for different releases for separate files If the files have test cases with unique names
Reports and Cyclone	After executing a few test cases that are failed and a few from a different status.	The hot test cases report statistic for most failed test cases and most executed test cases does not reflect immediately.	Most executed and most failed report job is run every day mid night as part of a quartz job and collects the data. The data is not real time and cannot be updated until the job is run on its' scheduled time. This is an ETL job and can be viewed in the database by name "ZpadCacheJob" under Qrtz_Cron_Trigger table, the cron expression can be modified if it is required to change its' default execution time. To reflect the results of the report for most executed and failed test cases, users need to fire the clearcache API call. Fire the cacheclear API call. GET: http://localhost:80/flex/services/rest/v3/zpad/clearcache Restart the service.
Dashboards	While using the Test Execution by Cycle gadget and after drilling down in the gadget in the dashboard on IE 11.	While navigating back to the gadget/dashboard after drilling down the Test Execution by Cycle graph, the graph data tends to overlap.	We recommend using Google Chrome as the suggested browser as the Zephyr framework works best on Chrome. Chrome also allows users to download the Zephyr Exploratory chrome extension plugin.
Dashboards	When using the daily pulse gadget in the dashboard for a selected date range that ends in a date (IE. yesterday's date).	The Testcase, Execution, and Defect count and chart will show more than the end date selected if there's data past the end date.	Users shouldn't select an end date that is in the past. The end date should always be in the future to allow the gadget to keep viewing the daily data for test cases created, executed, and defects linked.
Dashboards	When using the daily pulse gadget in the dashboard for a selected custom date range from a previous date.	The Testcase, Execution, and Defect count will not display the live results and will show the count of the end date selected from the custom date range.	Suppose we have a selected date range for 7 days. If we have data for 5 days and the data within the last 2 days is not there, then the last day trend data will reflect the count. The count and chart will show data for the 6th day because we have the trend data.