April 2022 Release

This release covers:

New:

• WAS FP21 support

Fix:

• Helm version detection fix
• Documentation updates

March 2022 Release

Fix:

1. Fix an issue when deploying with Oracle backend where DB2 tasks are run
2. Address issue in setup-kubectl in cluster environment when copying files
3. Set Invite user type default to be external
4. Fix Huddo Boards deployment issues
5. Remove redundant was-dmgr-config-interceptor folder

Improvement

1. Stop setup-kubernetes.yml playbook if one of the roles fails.
2. Use --delete-emptydir-data in Kubernetes drain node for later version
3. New enable_action_center variable to enable/disable the Action Center
4. Documentation updates

Feb 2022 Release

Fix:

1. Fix for updateInstaller extracted location Issue #165
2. Remove disable fingerprinting Issue #184
3. Use db2jcc4.jar Issue #186
4. Enable QuickResults by default for Connections 6.5 CR1
5. Added any_errors_fatal for DB2 and NFS execution, removed ansible_distribution conditions. Issue #183

Improvement:

1. Automate Install/Uninstall tiny editors
2. Add task to install WebSphere Log4j interim fix PH42762
3. Add task to configure mail server and port by setting mail_outgoing_server and mail_smtp_port
4. Add TDI cleanup to avoid multiple SLF4J bindings issue
5. Added timeout of 10 min for ES7 and OrientMe.
6. Docker change: Since Go version 1.15, the deprecated, legacy behavior of treating the CommonName field on X.509 certificates as a hostname when no Subject Alternative Names (SAN) are present is now disabled by default. A new valid certificate is created to include the subjectAltName property, and added directly when creating an SSL self-signed certificate using openssl command.

Change:

1. Upgrade default versions for infra:

• Docker v20.10.12
• Kubernetes v1.21.7
• Helm v3.7.2
• HAProxy v2.5.1
• Containerd v1.4.12
• IBM DB2 v11.5.6
• WebSphere FP20

Jan 2022 Release

Fix:

• Download to PDF failed for Docs due to missing libraries
• Fixed start_cluster/stop_cluster failure due to previous stop/start cluster process is still in progress.

Improvement:

• New variables documentation page
• New playbook to set global moderator after installation.
• Add steps to force traffic to use TLS1.2
• Update Outlook Add-in deployment using CNX Ingress to work with Outlook Add-in version 21.01
• Cleanup playbook improvements. Added a new flag preserve_nfs_data. Data will be preserved/deleted according to the flag value.
• Point installation to an existing database #132
  - Added new inventory files for DB2, Oracle and MSSQL under environments/examples/existing_database
  - Added a new playbook to point installation to an existing DB at playbooks/hcl/connections-existing-database.yml. This playbook verifies if all mandatory variables are set upfront in the inventory file. It fails if any of the mandatory variable is missing.
  - Adding a new flag install_latest_db2 in the inventory. If the flag is 'true' then DB2 v11.5.6 will be installed (when install kit is available). If the flag is 'false' or missing then DB2 v11.1 will be installed.
• New var tdi_sol_location to specify a different tdisol tar for SDI JRE8
• WAS FP20 install support
• Support for RHEL8

Change:

• Language selector will be disabled by default, use cnx_enable_lang_selector var to enable the feature.

Oct 2021 Release

• Standalone playbook to delete WAS temp dir to clear cache after Connections upgrade
• Standalone playbook to run ES WAS configuration if needed after deployment
• Azure AD OIDC setup (not fully tested)
• Docs pre-install vars check, fixed issue where "Download to PDF" does not work due to missing libraries
• Standalone playbook to clean up existing connections environment which removes WAS, IHS, Kubernetes, DB2/Oracle, NFS data, and Containerd images
• Standalone playbook to unmount NFS folders and clean up NFS data
• Standalone playbook to delete Containerd images

Maintenance and RHEL8 support

This release provides the updates in automation needed to support it properly on RHEL 8 (it was tested on the latest RedHat Enterprise Linux 8.4).

• Oracle 19c installation now works on RHEL8
• OpenLDAP build added that works on RHEL8 as well
• Docker, containerd.io and Kubernetes install updated to work with RHEL8.
• Updated dependencies for WASND and Connections to support RHEL8.

Other fixes and improvements:

• WASND upgraded to 8.5.5.19. The default WAS ND version from this release on is 8.5.5.19.
• SOAP timeout fixed on multiple places.
• TDI JDK upgrade fixed (would be needed with Connections 8).
• Default Haproxy version bumped to latest and greatest.
• Default Helm version bumped to latest and greatest.
• Default Kubernetes version bumped to 1.19.11

Known issues

• IBM IHS is having some issues with mod_ibm_ssl.so on all versions of RHEL8 we tested it with (from 8.0 to 8.4). We suggest to, in this case, either contact IBM support, or use RHEL7 for IHS only (supported by automation as well).
• HCL Connections Docs wasn't tested because the step creating jobs on WAS is failing internally on WAS on RHEL8.

0.4.2 Bugfix/minor feature release

• Allow customising Component Pack download/extraction folder
• Replace synchronise with copy module when handling ES certificates
• If Kudosboards is not enabled don't set it in httpd.conf
• Cleaning up K8s PV NFS content on cleanup
• Restarting Docs clusters when restarting everything, getting vars from NFS servers on cleanup
• Minor MT fix
• Handling pauses/waits better for Component Pack components
• Replace pause for Docker Registry with proper service check
• Adding retry if DMGR didn't manage to fully recover
• Fixing old Calico versions handling
• Update Kubernetes init when not using LB in front of it

0.4.1 Bugfix release

This release contains minor bug and feature fixes:

• Disabled explicit Samesite=None in httpd.conf
• Don't force swap creation if swap is already enabled and used
• Don't use explicit pauses in Component pack, speed up the execution
• Ensure specific kubectl version is installed
• Bump default Kubernetes version to 1.18.19
• Bump default Helm version to 3.5.4
• Couple of bug fixes related to HCL Connections Docs installation

April 2021 Release

This release contains the major refactoring and is backwards compatible with the previous releases.

The major topics of the current release are:

• The way how variables are used is refactored, and we tried to deduplicate it as much as possible. We tried to lower the number of manual changes as much as possible. See quick start or any of other examples to get the idea of what was done.
• HCL Connections deployment topologies are now controlled through the single variable (small/medium/large deployment)

Minor fixes:

• Auto scaling disabled for 6.5.* to avoid breaking ElasticSearch
• Testing latest versions of Kubernetes on branches 1.18, 1.19 and 1.20. Bumped base version for Kubernetes to 1.18.18 and Docker to 19.03.15
• Made Kubernetes deployment easier by removing variable specifying specifically if the deployment is single node or multiple nodes.
• Prometheus Exporter moved outside of Component Pack setup.
• Added Prometheus Node Exporter support.
• Added Prometheus Oracle Exporter support.
• Added support for syncing LTPA timeout between Connections and Component Pack.
• DB2 module refactoring - skip steps if DB2 is already up and running to save time.
• Adding Java 8 support for TDI for new versions of Connections (not yet officially released and supported)
• Adding Oracle cleanup support
• Fixing cleanup scripts
• Fixing ingress controller wait as per https://kubernetes.github.io/ingress-nginx/deploy/
• Fixing bootstrap pod cleanup on reinstalls
• Fixing Quick Start docs and replacing examples.

Cumulative update March 23rd 2021

Cumulative updates - March 2021

Major update - adding support for Microsoft SQL Server 2019

This release introduces new concept for managing JDBC for different database backends. Please see this howto.
This can be considered a breaking change since it introduces the differences with the previous concept.
This is also the first release which supports Microsoft SQL Server 2019.

Documentation

• Updated README.md
• Added new how to for supporting different database backends.

Connections, Component Pack and Docs fixes:

• Removing Prometheus from Component Pack, moving it to own module
• Enforcing replica count for Component Pack and changing when scaling happens
• Adding support for MSSQL 2019 for Docs and Connections
• MT fixes

Third-party fixes:

• Fixing Docker support on CentOS 8.3
• Fixing Helm Tiller for Helm v2 for backwards compatibility
• Adding Prometheus Operator and Node Exporter modules
• Haproxy should point to Kubernetes workers not masters
• Adding support for MSSQL 2019
• Breaking change: adding switch for installing DB2/Oracle/MSSQL JDBC files. See new howto.
• Fixing (re)adding certificates on WAS
• Updating Kubernetes module, supporting 1.20+ more nativelly with new syntax
• Adding support for Kubernetes 1.18.16
• Refactoring OpenLDAP module
• Refactoring NFS module, supporting scaling of pods

Updating playbooks:

• Adding support for Oracle & MSSQL
• Adding playbooks for PromOperator and Node Exporter
• Fixing Connections Vagrant playbook