Changes in helm chart from version 1.0.x to 1.1.x
Changes to overall installation
Changes to Installation Types:
- APIs have been updated from StatefulSets to Deployments to reflect their stateless nature, which is now applied in the Helm chart.
Other
- The experimental "Fail-Over" feature, which was never used in production, has been removed.
- Load balancer configuration has been added.
- New GUI added TMS-GUI
Changes to values file
Updates made to values.yaml
file. This indicates new configurations that can be done.
Non-breaking Changes
Exposed existing configuration
Changes already existed in the dockers but were not exposed in helm chart
-
Exposed Connection Pool Configuration (
mysql.pool
): Added with defaults:size
:5
maxOverflow
:2
timeout
:3
recycle
:1800
-
Exposed Tenant Management Service (TMS) API (
tmsApi.host
andtmsApi.database
): Added with defaults:host
:"be-tms-svc"
database
:"tms"
-
Exposed System Events and Cache Settings: Added with defaults:
config.systemEvents.enabled
:"yes"
config.systemEvents.monitorPropertyChange
:null
config.systemEvents.collectVisitEnded
:"no"
config.cache.keepProfileInCacheFor
:3600
config.cache.keepSessionInCacheFor
:1800
-
Exposed APM, Experimental Features, and Additional Flags (
config.apm
): Added with defaults:apm.identificationEventProperty
:""
apm.tipType
:""
apm.eventType
:""
eff.enableLateProfileBinding
:"no"
features.enableEventDestinations
:"yes"
features.enableProfileDestinations
:"yes"
features.enableAudiences
:"yes"
features.enableWorkflow
:"yes"
features.enableEventValidation
:"yes"
features.enableEventReshaping
:"yes"
features.enableEventMapping
:"yes"
features.enableEventToProfileMapping
:"yes"
features.enableDataCompliance
:"yes"
features.enableEventSourceCheck
:"yes"
features.enableIdentificationPoints
:"yes"
-
Exposed Resource Limits and Requests: Added with defaults for memory and CPU allocations:
api.private.resources.limits.memory
:1000Mi
api.private.resources.limits.cpu
:500m
api.private.resources.requests.memory
:100Mi
api.private.resources.requests.cpu
:100m
api.public.resources.limits.memory
:300Mi
api.public.resources.limits.cpu
:500m
api.public.resources.requests.memory
:100Mi
api.public.resources.requests.cpu
:100m
worker.background.resources.limits.memory
:500Mi
worker.background.resources.limits.cpu
:1000m
worker.background.resources.requests.memory
:300Mi
worker.background.resources.requests.cpu
:300m
worker.collector.resources.limits.memory
:500Mi
worker.collector.resources.limits.cpu
:1000m
worker.collector.resources.requests.memory
:300Mi
worker.collector.resources.requests.cpu
:300m
-
Exposed Worker Configuration (
worker.background
): New settings with defaults:worker.background.enabled
:true
worker.background.config.loggingLevel
:"INFO"
worker.background.config.bulker.bufferInactivityTimeOut
:10000
Updated Changes
- Version Tags: Updated from
1.0.0
to1.1.x
across services:api.image.tag
worker.background.image.tag
gui.image.tag
bridge.queue.image.tag
Removed Changes
- Telemetry and Digital Ocean Sections: Marked as not used (commented out). This was experimental feature.
- Coping Worker Configuration: Removed the configuration for
copingWorker
. No used anymore. - Removed Fail-over: Not used anymore
storage.failOver.enabled
:false
storage.failOver.size
:"1Gi"
Newly Added Changes
Changes in helm chart due to new functions
-
Node Affinity Configuration Examples: Added examples for affinity settings across services (e.g.,
api.private.nodeAffinity
,tmsGui.nodeAffinity
). Commented by default. See distributing load on different servers. -
Load Balancer Settings: Added across services with a default:
loadBalancer
:false
(inapi.private.service
,api.public.service
,tms.service
, andtmsGui.service
).
-
Collector and Workflow Configuration: New configurations with defaults:
- Collector settings (
worker.collector
):worker.collector.enabled
:false
worker.collector.config.loggingLevel
:"INFO"
worker.collector.config.bulker.bufferInactivityTimeOut
:10000
- Workflow settings (
worker.workflow
):worker.workflow.enabled
:false
worker.workflow.config.loggingLevel
:"INFO"
worker.workflow.config.bulker.bufferInactivityTimeOut
:10000
- Collector settings (
Changes that need attention
-
Load Balancers Disabled by Default: In the new Helm chart, load balancers are set to
false
by default. If you need to enable load balancers for any API, please modify your localvalues.yaml
file.- Enabling Load Balancers: To enable load balancers, update the
loadBalancer
setting totrue
in the following paths within yourvalues.yaml
:
- Enabling Load Balancers: To enable load balancers, update the
-
System Events Configuration Adjusted:
- Move
config.systemEvents
toconfig.systemEvents.enabled
if your local value files hasconfig.systemEvents
. - Move config
config.enableVisitEnded
toconfig.systemEvents.collectVisitEnded
if your local values file has this setting.
- Move
Make these changes in each relevant section to ensure load balancers are activated where needed.