Releases: merative/spm-kubernetes
Releases · merative/spm-kubernetes
[SPM v8] August Release 2021
v21.8.0
Breaking Changes
Ingress
andIngressClass
resources have graduated tonetworking.k8s.io/v1
, see Ingress graduates to General Availability- Due to ingress graduating to
networking.k8s.io/v1
spmingress.yaml
has been updated to checknetworking.k8s.io
version - Definition of the ingress path has been moved to the
spm.ingress.item
template
- Due to ingress graduating to
- SPM-Kubernetes default branch has been renamed from
master
tomain
Added
- Configure
PodMonitor
resources for theapps
producer and consumers pods and for themqserver
metrics pods to integrate with OpenShift's built-in Prometheus or Prometheus Operator
Changed
- The following helm-charts have been updated to chart version
21.8.0
:apps
,batch
,mqserver
,spm
,uawebapp
,web
,xmlserver
- Changed DB2 datasources isolation level for the
apps
producer and consumers pods. See Transaction control/Underlying design/DB2/Repeatable Read - Updated IBM Documentation link to SPM V8
- Clarified prerequisite software statements (#83)
Fixed
- Update Oracle Database driver name to
ojdbc8.jar
(#84) - Fixed issue where MQ pods deployed by MQ Operator on Openshift were not respecting tuning params
[SPM v7] August Release 2021
v21.6.2
Breaking Changes
- Following the upgrade to IBM MQ 9.2 in release
v21.6.1
, the below list of breaking changes has been introduced:- On existing deployments, the upgrade of IBM MQ requires the applications pods to be scaled down completely, followed by a delay to process any remaining messages, before updating the deployment
- Moved
global.mq.multiInstance.availabilityType
toglobal.mq.availabilityType
- Moved
global.mq.multiInstance.storageType
toglobal.mq.storageType
- Moved
global.mq.multiInstance.storageClassName
toglobal.mq.storageClassName
- Removed
global.mq.multiInstance
as a configuration, and the following children parameters:operatorsEnabled
,cephEnabled
,useDynamicProvisioning
,nfsEnabled
,nfsIP
,nfsFolder
,nfsMountOptions
- Support for manual creation of persistent volumes and persistent volume claims has been removed
- Removed
global.apps.common.persistence.subDir
as a configuration- The paths on the storage volume have now change from
<root>/<HelmRelease_or_subDir>/<podName>
to<root>/<podName>
. This is to accommodate pod initialization on start-up
- The paths on the storage volume have now change from
Ingress
andIngressClass
resources have graduated tonetworking.k8s.io/v1
, see Ingress graduates to General Availability- Due to ingress graduating to
networking.k8s.io/v1
spmingress.yaml
has been update to checknetworking.k8s.io
version. - Definition of the ingress path has been moved to the
spm.ingress.item
template
- Due to ingress graduating to
Added
- Configure
PodMonitor
resources for theapps
producer and consumers pods and for themqserver
metrics pods to integrate with OpenShift's built-in Prometheus or Prometheus Operator
Changed
- The following helm-charts have been updated to chart version
21.6.2
:apps
,batch
,mqserver
,spm
,uawebapp
,web
,xmlserver
. - Changed DB2 datasources isolation level for the
apps
producer and consumers pods. See Transaction control/Underlying design/DB2/Repeatable Read
Fixed
- Fixed issue where tuning params for resources on MQ pods created by the MQ operator were not being honoured
[SPM v8] July Release 2021
v21.7.0
This release introduces support for SPM 8.0.0.0 and contains breaking changes for deployment of SPM 7.0.11.0
If you are deploying a version of SPM prior to 8.0.0.0, see therelease/v7
branch (or21.6.1
tag)
Breaking Changes
- Due to the upgrade to IBM MQ 9.2, the following breaking changes have been introduced:
- On existing deployments, the upgrade of IBM MQ requires the applications pods to be scaled down completely, followed by a delay to process any remaining messages, before updating the deployment
- Moved
global.mq.multiInstance.availabilityType
toglobal.mq.availabilityType
- Moved
global.mq.multiInstance.storageType
toglobal.mq.storageType
- Moved
global.mq.multiInstance.storageClassName
toglobal.mq.storageClassName
- Removed
global.mq.multiInstance
as a configuration, and the following children parameters:operatorsEnabled
,cephEnabled
,useDynamicProvisioning
,nfsEnabled
,nfsIP
,nfsFolder
,nfsMountOptions
- Support for manual creation of persistent volumes and persistent volume claims has been removed
- Removed
global.apps.common.persistence.subDir
as a configuration- The paths on the storage volume have now change from
<root>/<HelmRelease_or_subDir>/<podName>
to<root>/<podName>
. This is to accommodate pod initialization on start-up
- The paths on the storage volume have now change from
Added
- Add TLS certificates for Secure XML server
- Add OpenID configuration for authentication using JSON Web Tokens for SPM Chatbot only
- Add signing certificates for JSON Web Tokens for SPM Chatbot only
- Added SPM 8.0.0.0 annotations to SPM v8 specific content in the runbook
- Added known issue to account for WebSphere Liberty timeout messages WTRN0006W and WTRN0124I
- Add the ability to tune Java options for the XMLServer application
Changed
- Upgraded IBM MQ image from
9.1.5
to9.2.2
- Updated IBM MQ Resource Adapter to
9.2.2.0
- Starting from
spm-kubernetes
release21.7.0
, chart versions have been updated to align with the release version.- The following helm-charts have been updated:
apps
,batch
,mqserver
,spm
,uawebapp
,web
,xmlserver
. - All changes in helm-charts will now detailed in the
CHANGELOG.md
file
- The following helm-charts have been updated:
kubeVersion
for all helm-charts updated to">=1.19"
appVersion
for all helm-charts updated to8.0.0.0
- Liberty persistent timer tables are no longer created per pod, but are created and shared by pod-type
- Updated runbook pre-requisites page to add information on Docker
- Changed MQ configuration for the
apps
producer and consumers pods to be using separated channels - Moved tuning settings from
initContainer
to a new ConfigMap - IBM Documentation has now replaced IBM Knowledge Center. Runbook links have been updated accordingly.
Fixed
- Fixed XML Server shutdown by adding a pod preStop hook
Removed
- Removed
RELEASENOTES.md
for the following helm-charts:apps
,batch
,mqserver
,spm
,uawebapp
,web
,xmlserver
.- All changes in helm-charts will now detailed in the
CHANGELOG.md
file
- All changes in helm-charts will now detailed in the
- Removed
global.apps.common.persistence.subDir
as a configuration - Removed
persistence.subDir
from the following helm-charts:apps
,batch
,spm
,uawebapp
,web
,xmlserver
- Removed obsolete Helm Chart known issue
[SPM v7] July Release 2021
v21.6.1
Breaking Changes
- Removed
global.apps.common.persistence.subDir
as a configuration- The paths on the storage volume have now change from
<root>/<HelmRelease_or_subDir>/<podName>
to<root>/<podName>
. This is to accommodate pod initialization on start-up
- The paths on the storage volume have now change from
Added
- Add the ability to tune java options for the XMLServer application
- Added known issue to account for WebSphere Liberty timeout messages WTRN0006W and WTRN0124I
Changed
- Upgraded IBM MQ image from
9.1.5
to9.2.2
- Updated IBM MQ Resource Adapter to
9.2.2.0
- Updated runbook pre-requisites page to add information on Docker.
- Changed MQ configuration for the
apps
producer and consumers pods to be using separated channels - Starting from
spm-kubernetes
release21.6.1
, chart versions have been updated to align with the release version.- The following helm-charts have been updated:
apps
,batch
,mqserver
,spm
,uawebapp
,web
,xmlserver
. - All changes in helm-charts will now detailed in the
CHANGELOG.md
file
- The following helm-charts have been updated:
kubeVersion
for all helm-charts updated to">=1.19"
- Moved tuning settings from
initContainer
to a new ConfigMap - IBM Documentation has now replaced IBM Knowledge Center. Runbook links have been updated accordingly.
Fixed
- Fixed XML Server shutdown by adding a pod
preStop
hook
Removed
- Removed
RELEASENOTES.md
for the following helm-charts:apps
,batch
,mqserver
,spm
,uawebapp
,web
,xmlserver
.- All changes in helm-charts will now detailed in the
CHANGELOG.md
file
- All changes in helm-charts will now detailed in the
- Removed
global.apps.common.persistence.subDir
as a configuration - Removed
persistence.subDir
from the following helm-charts:apps
,batch
,spm
,uawebapp
,web
,xmlserver
- Removed obsolete Helm Chart known issue
June Release 2021
v21.6.0
Added
- Added capability to tune Kubernetes resources for MQ pods for individual applications
- Added clarification regarding sample values files.
- Added JVM garbage collection and tuning settings for XML server pods
- Updated sample override files to include example settings
- Added capability to specify thread pool size, thread queue size, and socket timeout value for XML server pods
- Added capability to tune various JMS tuning parameters for producers and consumers
- Extended jmsActivationSpec to include maxEndpoints parameter for the following JMS queues:
- DPError
- WorkflowError
- CuramDeadMessageQueue
- DPEnactment
- WorkflowEnactment
- WorkflowActivity
- Extended jmsActivationSpec to include maxEndpoints parameter for the following JMS queues:
- Added capability to tune the minPoolSize setting of the JMS connection manager
- Documented how to tune the resources for the uawebapp, web and xmlserver pods
Changed
- Upgrade to latest Gatsby library
- Updated the minimum supported version of Kubernetes to
1.19
Removed
- Removed support for Kubernetes
1.18
May 2021 Release fix drop 1
May 2021 Release
v21.5.0
Added
- Added capability for Prometheus to scrape Liberty metrics from pods
- Added clarification in the Runbook introduction page regarding the flexibility to modify or develop Helm Charts or Dockerfiles
- Added information in the MustGather introduction page about Helm Charts and Dockerfiles
- Added capability to tune various database and JMS tuning parameters for individual producers and consumers
- Created tuning-values.yaml as example.
- Added capability for Docker Hub details to be used to avoid toomanyrequests (#69)
- Added capability to set applications properties at deployment
Changed
- Allows for the simultaneous scraping of multiple metrics sources (eg JMX and Liberty) on
apps
charts - Moved
HTTPSessionDatabase
default configuration to timer based - Update
unzip
when unpacking the client EAR file - Ensure MQ directory structure exists, when using NFS (#31)
- Upgrade to Gatsby v3 (and associated dependencies)
April 2021 Release fix drop 1
IBM MQ XAER_PROTO issue has been observed in the JMS Consumer pods when JMS transactions is processed by a newly deployed pod, at the initial JMS transactions. An interim release has been created to clarify as a known issue.
v21.4.1
Added
- IBM MQ XAER_PROTO reported as a known issue (closes #73 )
April 2021 Release
v21.4.0
Added
- Extended WebSphere Liberty tuning options
Changed
- Changed the OpenShift reference architecture diagram to provide additional clarity regarding the statefulsets
- Improved pod labels for compatibility with service meshes (#61)
- Improved navigation links to avoid 404 errors (#68)
Removed
- Removed the
wait-for-database
initContainer from producer and consumer pods- Validation of the database configuration is already handled by the pre-install hook
March 2021 Release
v21.3.0
Added
- Adds values from
podAnnotations
at deployment ofmqserver
chart - Pass parameter to MQ Operators to enable or disable the metrics
Changed
- Clarify instructions for building base images (closes #63)
- Clarify initial setup of CodeReady Containers
- Update recommended drivers for running Minikube (closes #59)
Fixed
- Remove references to a non-existent pull secret for service accounts