The new version of Polarion is available  –  Polarion ALM version 18.3 (Service Release 3).

It is the last service release before our upcoming major release in March 2019. As usual, this last service release contains more quality improvements than big features.

We concentrated on performance improvements and technology update, based on feedback from our customers who use our solutions every day, as well as ongoing quality improvements.

Significant enhancements were done to the following:

• Teamcenter-Polarion integration – updating files in Teamcenter and new workflow function


• OpenJDK 11 support – OpenJDK 11 supported as server JVM


• Scalability on large data – significant scalability boost and reduced memory consumption 

Teamcenter-Polarion Integration

This release comes with new version 4.1 of Teamcenter-Polarion integration that adds support for Teamcenter version 11.6 and 12.1 together with Active Workspace 4.1.

Updating files in Teamcenter

This release also enhances support for handling of attachments, providing the possibility to add or update files in existing Teamcenter datasets.

The file upload performance has been improved, especially when uploading multiple large files.

Update file in Teamcenter

Easy to configure CreateRemoteObject workflow function

To make the process orchestration and traceability between Polarion and Teamcenter even more easy for the admins, a new Polarion workflow function that creates objects in Teamcenter has been introduced. And its configuration has been significantly simplified. This function uses OSLC for connecting to Teamcenter to create a new object in a much faster and reliable way.

Following is an illustration of the only mandatory parameters, while the parameters for setting string properties are still supported with the same syntax as original and now deprecated CreateTeamcenterObject workflow function.

Configuration of CreateRemoteObject workflow function
The connection to Teamcenter is now established automatically using the friend server configuration, so users do not have to provide any additional connection configuration if a common single sign-on via Teamcenter SSO is configured for both products.

Support of OpenJDK 11

Polarion catches up with the most recent version of JVM and becomes fully compatible with Java 11, relying on and recommending the OpenJDK 11 distribution. Polarion 18.3 still supports Oracle JDK 8, but this support will be discontinued starting with the next Polarion release (Polarion 19, scheduled for release in the spring of 2019).

We recommend switching to OpenJDK 11 with Polarion 18.3, to allow the time to accommodate in case any custom Polarion extensions need to be updated, and also to gain performance improvements offered by running Polarion on the new JVM.

Detailed information on updating the Java runtime can be found in the Third-party.txt and the Install Guide PDF files for the corresponding platform, all bundled in the Polarion update distribution.

Performance boost on Java 11

New versions of Java sometimes do provide more optimized processing that results in a minor speedup of Polarion’s use cases. The transition from the previously supported Java 8 to up to date Java 11 did not have a significant impact on the majority of use cases, typically speeding them up in terms of single percent. However, there a few scenarios that show a more substantial impact, especially when it comes to processing of big HTML content.

Scenario	Relative improvement
Document compare in LiveDocs	26% faster
Work Items compare in LiveDocs	10% faster
Opening cached LiveDoc	13% faster

Other Enhancements

Robustness of the DOORS connector

The robustness of the DOORS connector that was introduced in Polarion 18.1 has been improved as we work with customers that migrate DOORS modules to Polarion LiveDocs at enterprise scale.

Connectivity issues with DOORS are detected faster to prevent the import hanging for a long time if the DOORS client crashes or if it’s stuck.

Attributes from DOORS that cannot be auto-configured are now reported in the mapping UI to bring these problems to administrator’s attention quickly.

Several issues that resulted in the crashing of the import were fixed:

• DPP-174598 – Doors import crashes on dead links


• DPP-175258 – Inefficient loading of links during DOORS synchronization


• DPP-163121 – Importing additional rich-text attributes will fail ReqIF Import

Updated Eclipse platform

Polarion 18.3 also features an updated Eclipse platform — version 4.8, a.k.a. “Photon”. This was a necessary step to enable Java 11 support, but it also brings several enhancements and performance improvements of its own.

Authors of Polarion extensions and plug-ins should consult the Configuration.txt file, which is bundled in the Polarion update distribution and discusses the impact of the new Eclipse platform on custom Polarion plug-ins.

Updated online Help

A notable user-visible improvement brought by the Eclipse update is a new Help platform, that brings the following enhancements:

1. Breadcrumb navigation


2. “Link with Contents” automatically synchronizes TOC selection when following hyperlinks


3. The polished visual style, especially for the Help landing page


4. Enhanced Print, Home and Search functionality


5. Possibility to refer to anchors within a help page via URL (easy copy from TOC or links)


6. Security fixes

 Enhanced on-line Help

Upgrade of the Tomcat to version 9

Polarion leverages Apache Tomcat as one of its core technologies for authentication and servlet deployment. The previously used Tomcat 8.0 has been upgraded to Tomcat 9.0.12 to leverage latest security fixes as well as functional and performance enhancements.

For more details about changes please refer to the changelogs of Tomcat incremental versions:
https://tomcat.apache.org/tomcat-8.5-doc/changelog.html 
https://tomcat.apache.org/tomcat-9.0-doc/changelog.html 

Support Office 2019

Microsoft Office 2019 was tested and verified by our QA team to that ensure round-trip and other scenarios are working with the new version of the Microsoft Office suite.

Support for Firefox 63

Support for the latest Firefox 63 was tested and verified by our QA team.

Performance and Scalability

Tested with 1.000 users and 5.000.000 Work Items

The amount of data Polarion is processing is growing with every customer as our customers expand their Polarion usage by:

• on-boarding additional teams


• expanding to different divisions


• covering a larger part of the business processes


• starting new projects and products


• adoption of test automation approaches to accelerate the development

Companies are growing bigger and existing enterprise customers with big data experience exponential data growth. The question of scalability and support of large data sets becomes more and more critical.

The scalability testing on large data conducted by Polarion’s R&D team helps us to assess scalability limits and find potential bottlenecks and critical parts in the Polarion extensibility and scalability. Our team has been building special parallelized high load test scenarios based on typical Polarion user roles (like Manager, Developer, Tester…) simulating the daily usage of the product in large enterprises.

With Polarion 18.3 our team was able to successfully pass load tests simulating 1.000 concurrent users working with the system on a repository that contains 5.000.000 Work Items in 300 projects. All that on relatively small virtualized environment composed of 5 machines:

• 3 Polarion nodes – 4 CPU, 64 GB RAM


• coordinator – 2 CPU, 4 GB RAM


• shared storage – 4 CPU, 64 GB RAM

The chart below illustrates the distribution of different business roles in our test as they execute through their typical flow for 8 hours at a scale of 1000 users.

Distribution of personas
Each individual operation (e.g. save LiveDoc) defines its threshold limit duration that must be satisfied by at least 90% of the executions. The graph below shows the stability of response times for individual operations during the test.

Response times of individual scenarios

Optimization of Lucene queries with sorting on large data

One of the major optimizations delivered in this release was identified by our scalability tests on very large data. In previous versions, the Lucene index search with sorting has allocated a large amount of memory for every search. Consequently, it has caused high memory consumption for each request and a lot of work for Garbage Collector afterward.

However, the vast majority of Lucene search operations return only a relatively few results, so the approach was wasteful of memory resources. The optimization improves the search operation with sorting so that it allocates a much smaller amount of memory at the outset, and only if the result set is bigger do we allocate more. This potentially saves megabytes of memory for every query on large data, which directly translates into higher concurrent throughput of the system.

Optimized evaluation for Subversion-level access permissions

Subversion provides means for granular control of access to each location per user’s role. This permission control ensures that even users with direct Subversion access do not see any data they are not allowed to see. This Subversion access check is performed for every Polarion operation as well, ensuring that users can only see data which they are allowed to see both by Polarion permissions and Subversion-level access permissions.

The previous versions of Polarion could suffer from lock contention in case of slow network connection among cluster nodes and the shared storage. Or the lock could be acquired for a very long time by specific operations, e.g. when triggering the action “Synchronize SVN Access File” in Roles administration. While this action takes just a few moments for small repositories, it can take minutes on larger data repositories.

The new Polarion version reduces the file lock impact and also optimizes the access file synchronization process itself, making it non-blocking, memory modest and significantly faster. Measurements on our scalability tests repository show improvement from 1 hour to 15 seconds.

Notable Issue Fixes

• DPP-107583 – Separated by – Pie and Trend Chart widgets do not render the correct sections when rendering Work Items separated by type-specific enumerations


• DPP-165165 – Description read-only field pops up in a LiveDoc’s Work Item Properties sidebar when “Description” is a required field


• DPP-167871 – The correct ordering and structure is not maintained when Polarion XML is imported into an MS project


• DPP-168571 – Regression in 3.17.1: Work Items cannot be created in the DLE if the workflow requires the Description field in the initial action

Update Information

Version 18.3 is an update for all Polarion ALM products. It is free to all customers with a current maintenance subscription. You can download the update distribution at https://polarion.plm.automation.siemens.com/downloads/update. For details, see the bundled HOW_TO_INSTALL_THIS_UPDATE.txt in the update distribution package.

Evaluation

If you would like to evaluate this release before updating your production installation, simply visit https://polarion.plm.automation.siemens.com/downloads, download the product of your choice, install it and use the built-in 30-day evaluation license.

​​

If you have any questions or comments, please don’t hesitate to contact me or your Polarion technical support contact. On behalf of our entire team, thank you for using Polarion ALM.

Best regards,
Radek Krotil
Polarion ALM Product Management