[GitHub] cbickel closed pull request #3547: Remove bookkeeping setting which is no longer applicable.

[GitBox <gi...@apache.org>]

cbickel closed pull request #3547: Remove bookkeeping setting which is no longer applicable.
URL: https://github.com/apache/incubator-openwhisk/pull/3547
 
 
   

This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:

As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):

diff --git a/ansible/roles/controller/tasks/deploy.yml b/ansible/roles/controller/tasks/deploy.yml
index b9e8c788d2..87167f5041 100644
--- a/ansible/roles/controller/tasks/deploy.yml
+++ b/ansible/roles/controller/tasks/deploy.yml
@@ -171,7 +171,6 @@
       "CONFIG_whisk_runtimes_bypassPullForLocalImages": "{{ runtimes_bypass_pull_for_local_images | default() }}"
       "CONFIG_whisk_runtimes_localImagePrefix": "{{ runtimes_local_image_prefix | default() }}"
 
-      "CONTROLLER_LOCALBOOKKEEPING": "{{ controller.localBookkeeping }}"
       "AKKA_CLUSTER_SEED_NODES": "{{seed_nodes_list | join(' ') }}"
       "METRICS_KAMON": "{{ metrics.kamon.enabled }}"
       "METRICS_KAMON_TAGS": "{{ metrics.kamon.tags }}"
diff --git a/common/scala/src/main/scala/whisk/core/WhiskConfig.scala b/common/scala/src/main/scala/whisk/core/WhiskConfig.scala
index 87793230ef..5fc599fc21 100644
--- a/common/scala/src/main/scala/whisk/core/WhiskConfig.scala
+++ b/common/scala/src/main/scala/whisk/core/WhiskConfig.scala
@@ -87,7 +87,6 @@ class WhiskConfig(requiredProperties: Map[String, String],
   val actionInvokeSystemOverloadLimit = this(WhiskConfig.actionInvokeSystemOverloadLimit)
   val actionSequenceLimit = this(WhiskConfig.actionSequenceMaxLimit)
   val controllerSeedNodes = this(WhiskConfig.controllerSeedNodes)
-  val controllerLocalBookkeeping = getAsBoolean(WhiskConfig.controllerLocalBookkeeping, false)
 }
 
 object WhiskConfig {
@@ -209,7 +208,6 @@ object WhiskConfig {
   val actionInvokeSystemOverloadLimit = "limits.actions.invokes.concurrentInSystem"
   val triggerFirePerMinuteLimit = "limits.triggers.fires.perMinute"
   val controllerSeedNodes = "akka.cluster.seed.nodes"
-  val controllerLocalBookkeeping = "controller.localBookkeeping"
 }
 
 object ConfigKeys {
diff --git a/core/controller/src/main/scala/whisk/core/loadBalancer/ShardingContainerPoolBalancer.scala b/core/controller/src/main/scala/whisk/core/loadBalancer/ShardingContainerPoolBalancer.scala
index 718079ead6..d10f03e08e 100644
--- a/core/controller/src/main/scala/whisk/core/loadBalancer/ShardingContainerPoolBalancer.scala
+++ b/core/controller/src/main/scala/whisk/core/loadBalancer/ShardingContainerPoolBalancer.scala
@@ -305,9 +305,7 @@ object ShardingContainerPoolBalancer extends LoadBalancerProvider {
     logging: Logging,
     materializer: ActorMaterializer): LoadBalancer = new ShardingContainerPoolBalancer(whiskConfig, instance)
 
-  def requiredProperties: Map[String, String] =
-    kafkaHosts ++
-      Map(controllerLocalBookkeeping -> null, controllerSeedNodes -> null)
+  def requiredProperties: Map[String, String] = kafkaHosts ++ Map(controllerSeedNodes -> null)
 
   /** Generates a hash based on the string representation of namespace and action */
   def generateHash(namespace: EntityName, action: FullyQualifiedEntityName): Int = {
diff --git a/docs/deploy.md b/docs/deploy.md
index 49c2636758..1ad8456af5 100644
--- a/docs/deploy.md
+++ b/docs/deploy.md
@@ -2,7 +2,7 @@ This page documents configuration options that should be considered when deployi
 
 # Controller Clustering
 
-The system can be configured to use Akka clustering to manage the distributed state of the Contoller's load balancing algorithm.  This imposes the following constriaints on a deployment
+The system can be configured to use Akka clustering to manage the distributed state of the Contoller's load balancing algorithm.  This imposes the following constraints on a deployment
 
 
 ## Controller nodes must have static IPs/Port combination.
@@ -19,7 +19,7 @@ https://doc.akka.io/docs/akka/2.5.4/scala/cluster-usage.html
 
 ## Shared state vs. Sharding
 
-OpenWhisk supports both a shared state and a sharding model. By default the shared-state loadbalancer is used. The sharding loadbalancer is the newer implementation and scheduled to eventually supersede the shared-state implementation and become the default. To configure your system to use the sharding implementation, set `controller_loadbalancer_spi` to `whisk.core.loadBalancer.ShardingContainerPoolBalancer`.
+OpenWhisk used to support both shared state and a sharding model. The former has since been deprecated and removed.
 
 The sharding loadbalancer has the caveat of being limited in its scalability in its current implementation. It uses "horizontal" sharding, which means that the slots on each invoker are evenly divided to the loadbalancers. For example: In a system with 2 loadbalancers and invokers which have 16 slots each, each loadbalancer would get 8 slots on each invoker. In this specific case, a cluster of loadbalancers > 16 instances does not make sense, since each loadbalancer would only have a fraction of a slot above that. The code guards against that but it is strongly recommended not to deploy more sharding loadbalancers than there are slots on each invoker.
 


 

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services