[GitHub] csantanapr commented on issue #3558: WebAction async
csantanapr commented on issue #3558: WebAction async URL: https://github.com/apache/incubator-openwhisk/issues/3558#issuecomment-383249089 Yep a query and header parameter was brought up but this would need to be special named as we don’t want to conflict with a header or query intended for the action. A header @rabbah mentioned already is kind of pain to add a header in some tools like Travis only allows a url to be call On my last proposal I indicated that this issue is only to cover the minimum to do invoke an action in async way nothing to deal with triggers. /web/trigger can be discuss in another issue, right I see it as a way to make a trigger fire with no whisk authentication the trigger/rule/actions owner gets assigned the bill of the duration :-), allowing anything to fire the trigger. 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] csantanapr commented on issue #3558: WebAction async
csantanapr commented on issue #3558: WebAction async URL: https://github.com/apache/incubator-openwhisk/issues/3558#issuecomment-383249089 Yep a query and header parameter was brought up but this would need to be special named as we don’t want to conflict with a header or query intended for the action. On my last proposal I indicated that this issue is only to cover the minimum to do invoke an action in async way nothing to deal with triggers. /web/trigger can be discuss in another issue, right I see it as a way to make a trigger fire with no whisk authentication the trigger/rule/actions owner gets assigned the bill of the duration :-), allowing anything to fire the trigger. 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] csantanapr closed pull request #186: Helm chart improvements
csantanapr closed pull request #186: Helm chart improvements URL: https://github.com/apache/incubator-openwhisk-deploy-kube/pull/186 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/helm/README.md b/helm/README.md index 3624054..6389115 100644 --- a/helm/README.md +++ b/helm/README.md @@ -1,50 +1,26 @@ ## Overview -[Apache OpenWhisk](http://www.openwhisk.org) is a popular container-native serverless platform. -This chart targets to deploy an OpenWhisk cluster over Kubernetes cluster. +This directory contains a Helm chart can be used to deploy Apache OpenWhisk to a Kubernetes cluster. -This chart is designed to deploy the following 6 kinds of containers on a running Kubernetes cluster: +It currently does not support all of the options available in the more +manual deployment process described in the top-level README.md, +however we are actively working on closing the gaps. -- `zookeeper`: deployed in Kubernetes as a `Service`(backended by a `Deployment`). Provides `zookeeper` service for `kafka` node. -- `kafka`: deployed in Kubernetes as a `StatefulSet`. Depends on `zookeeper` service and provides message service for `controller` node and `invoker` node. -- `couchdb`: deployed in Kubernetes as a `Service`(backended by a `Deployment`). Provides database service for `controller` node and `invoker` node. -- `controller`: OpenWhisk controller component, deployed as a `StatefulSet`. -- `invoker`: OpenWhisk invoker component, deployed as a `StatefulSet`. -- `nginx`: deployed in Kubernetes as a combination of a `Service`(of type `NodePort`, backended by a `Deployment`), a `ConfigMap` and a `Secret`. Provides public access endpoint for end user to visit OpenWhisk. - -By default, deploying this chart will result in deploying 6 containers in a Kubernetes cluster, which forms a minimum deployment of a workable OpenWhisk environment. Please modify `values.yaml` and corresponding template files to deploy a bigger customized OpenWhisk cluster. +The default values used in the chart are designed to deploy a minimal +OpenWhisk deployment suitable for local development or testing on +minikube or a single node Kubernetes cluster. We will soon provide a +second set of default values suitable for larger-scale deployments. ## Deployment Steps Please follow the following steps in this section to use [Helm](https://github.com/kubernetes/helm) to deploy this chart. -### Step 1. Prepare Docker Images - -The first step is to prepare docker images used by this chart on your Kubernetes node. Here is a summary of the images needed: - -- `zookeeper`: uses `zookeeper:3.4` at present. -- `kafka`: uses `wurstmeister/kafka:0.11.0.1`. -- `couchdb`: uses `openwhisk/kube-couchdb`, this is a pre-built docker image made from official `openwhisk-on-kubernetes`. For more information, please check the [docker file](https://github.com/apache/incubator-openwhisk-deploy-kube/blob/master/docker/couchdb/Dockerfile) for this image. -- `controller`: uses OpenWhisk official `openwhisk/controller` image. -- `invoker`: uses OpenWhisk official `openwhisk/invoker` image. -- `nginx`: uses `nginx:1.11`. - -This chart provides default images for all of the above containers, so you can try deploy without building any images in advance. - -This chart will also automatically pull the following action images for `invoker` once deployment is finished: - -- openwhisk/nodejsactionbase -- openwhisk/nodejs6action -- openwhisk/dockerskeleton -- openwhisk/python2action -- openwhisk/python3action -- openwhisk/action-swift-v3.1.1 -- openwhisk/swift3action -- openwhisk/java8action +### Step 1. Prepare Kubernetes and Helm -### Step 2. Prepare Kubernetes and Helm +Make sure that you have a running Kubernetes cluster and a `kubectl` +client connected to this cluster as described in the [Requriements section](../README.md#requirements) of the main README.md. -Make sure that you have a running Kubernetes cluster and a `kubectl` client connected to this cluster. +### Step 2. Install and configure Helm Then please install [Helm](https://github.com/kubernetes/helm) and run the following command to init `Helm Tiller`: ```shell @@ -63,14 +39,26 @@ Then grant corresponding cluster role to `Helm` user: kubectl create clusterrolebinding tiller-cluster-admin --clusterrole=cluster-admin --serviceaccount=kube-system:default ``` -Finally, create a new namespace named `openwhisk`: +### Step 3. Prepare your cluster for OpenWhisk + +1. Create a new namespace named `openwhisk`: ```shell kubectl create namespace openwhisk ``` -Now, you should have an available environment to deploy OpenWhisk! +2. Identify the Kubernetes worker nodes that should be used to execute +user containers. Do this by labeling
[GitHub] dgrove-oss commented on issue #186: Helm chart improvements
dgrove-oss commented on issue #186: Helm chart improvements URL: https://github.com/apache/incubator-openwhisk-deploy-kube/pull/186#issuecomment-383238254 @csantanapr @rabbah -- would one of you be willing to review & merge if the changes look reasonable to you? Not doing everything the non-Helm install is doing yet, but assuming the docker images are already downloaded, it gets from nothing to a working basic install on minikube in about 3 minutes with just a couple of commands... 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] gorkem commented on issue #31: Java action that depends on packages jar file fails
gorkem commented on issue #31: Java action that depends on packages jar file fails URL: https://github.com/apache/incubator-openwhisk-runtime-java/issues/31#issuecomment-383222030 @rabbah Are dependencies packaged into /lib folder even supposed to work. This is not really a Java standard and I do not see anything to enable this on the runtime. @brunogirin Do your functions work OK when you package them as one jar? 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] dubee commented on issue #2855: Cache database attachments
dubee commented on issue #2855: Cache database attachments URL: https://github.com/apache/incubator-openwhisk/pull/2855#issuecomment-383188018 PG4 1600 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] dubee commented on issue #3416: Unit test PoolingRestClient
dubee commented on issue #3416: Unit test PoolingRestClient URL: https://github.com/apache/incubator-openwhisk/pull/3416#issuecomment-383187908 PG3 2154 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] dubee commented on issue #3416: Unit test PoolingRestClient
dubee commented on issue #3416: Unit test PoolingRestClient URL: https://github.com/apache/incubator-openwhisk/pull/3416#issuecomment-383187471 @markusthoemmes, I think that could be done in a different PR if we want to make requestQueue configurable. What do you think that advantages are there? 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rabbah commented on issue #3558: WebAction async
rabbah commented on issue #3558: WebAction async URL: https://github.com/apache/incubator-openwhisk/issues/3558#issuecomment-383142307 On /trigger my inclination is to support it only on webactions today (with an implicit trigger and rule - any webaction can be invoked in this way, asynchrinously, without creating a trigger or rule), but this will cause ambiguity later if we allow real triggers to be named so I think my suggestion was not well thought out at the time. 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rabbah commented on issue #3558: WebAction async
rabbah commented on issue #3558: WebAction async URL: https://github.com/apache/incubator-openwhisk/issues/3558#issuecomment-383141723 in a slack chat Carlos brought up the query parameter but we found that it could be ambiguous since it can be an parameter intended for the action itself. iirc he also suggested a privileged name such as __OW_ASYNC. 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ddragosd commented on issue #3558: WebAction async
ddragosd commented on issue #3558: WebAction async URL: https://github.com/apache/incubator-openwhisk/issues/3558#issuecomment-383140405 Just making sure we cover all options: was the query param `async=true` considered as well? On the `/trigger`: does the proposal suggest it should work with both “triggers” and “webactions”, or limit it to “webactions” only ? 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] ddragosd commented on issue #3199: One tid for the whole system.
ddragosd commented on issue #3199: One tid for the whole system. URL: https://github.com/apache/incubator-openwhisk/pull/3199#issuecomment-383134714 :+1: @rabbah 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] klcodanr commented on issue #3540: Vagrant Hello Not Starting on Mac
klcodanr commented on issue #3540: Vagrant Hello Not Starting on Mac URL: https://github.com/apache/incubator-openwhisk/issues/3540#issuecomment-383106148 Confirmed, thanks @csantanapr ! 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rabbah commented on a change in pull request #3563: Skip log collection if the log-limit is set to "0".
rabbah commented on a change in pull request #3563: Skip log collection if the
log-limit is set to "0".
URL:
https://github.com/apache/incubator-openwhisk/pull/3563#discussion_r183050508
##
File path: common/scala/src/main/resources/application.conf
##
@@ -156,6 +156,13 @@ whisk {
std = 256 m
}
+# action memory configuration
Review comment:
Wrong comment.
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:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] markusthoemmes opened a new pull request #3563: Skip log collection if the log-limit is set to "0".
markusthoemmes opened a new pull request #3563: Skip log collection if the log-limit is set to "0". URL: https://github.com/apache/incubator-openwhisk/pull/3563 Enables the per-action log-limit configuration to be read from config (min/max/std values respectively). It also skips the entire log-collecting process if that limit is set to 0 (no logs allowed). ## TODO - [ ] A test ## My changes affect the following components - [ ] API - [X] Controller - [ ] Message Bus (e.g., Kafka) - [ ] Loadbalancer - [X] Invoker - [ ] Intrinsic actions (e.g., sequences, conductors) - [ ] Data stores (e.g., CouchDB) - [ ] Tests - [ ] Deployment - [ ] CLI - [ ] General tooling - [ ] Documentation ## Types of changes - [ ] Bug fix (generally a non-breaking change which closes an issue). - [ ] Enhancement or new feature (adds new functionality). - [ ] Breaking change (a bug fix or enhancement which changes existing behavior). ## Checklist: - [X] I signed an [Apache CLA](https://github.com/apache/incubator-openwhisk/blob/master/CONTRIBUTING.md). - [X] I reviewed the [style guides](https://github.com/apache/incubator-openwhisk/wiki/Contributing:-Git-guidelines#code-readiness) and followed the recommendations (Travis CI will check :). - [ ] I added tests to cover my changes. - [ ] My changes require further changes to the documentation. - [ ] I updated the documentation where necessary. 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] rabbah commented on issue #3199: One tid for the whole system.
rabbah commented on issue #3199: One tid for the whole system. URL: https://github.com/apache/incubator-openwhisk/pull/3199#issuecomment-383076584 Makes sense @mdeuser - should use this as another opportunity to rip out another typed schema in the cli. Most values should be opaque and transient and their types inconsequential. 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] mdeuser commented on issue #3199: One tid for the whole system.
mdeuser commented on issue #3199: One tid for the whole system. URL: https://github.com/apache/incubator-openwhisk/pull/3199#issuecomment-383075407 @cbickel https://github.com/apache/incubator-openwhisk-client-go/blob/7671398729ab11fdaeb17e1f3ae308e3126f740a/whisk/client.go#L610 i think we should be careful of the sequence of the deployment of these changes. i.e. first release a version of the cli/plugin that handles the `code` value being either an integer or string. then deploy the backend. i'm not sure if there are other sdks that might be impacted by this api change... 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] cbickel commented on issue #3199: One tid for the whole system.
cbickel commented on issue #3199: One tid for the whole system. URL: https://github.com/apache/incubator-openwhisk/pull/3199#issuecomment-383070756 @mdeuser Can you point me to the correct position in the CLI where the integer for the code is required? Because I was not able to find it in the repositories https://github.com/apache/incubator-openwhisk-cli and https://github.com/apache/incubator-openwhisk-client-go 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] cbickel commented on a change in pull request #3199: One tid for the whole system.
cbickel commented on a change in pull request #3199: One tid for the whole
system.
URL:
https://github.com/apache/incubator-openwhisk/pull/3199#discussion_r182975152
##
File path: common/scala/src/main/scala/whisk/common/TransactionId.scala
##
@@ -200,58 +196,42 @@ object TransactionId {
val metricsKamonTags: Boolean =
sys.env.get("METRICS_KAMON_TAGS").getOrElse("False").toBoolean
val metricsLog: Boolean =
sys.env.get("METRICS_LOG").getOrElse("True").toBoolean
- val unknown = TransactionId(0)
- val testing = TransactionId(-1) // Common id for for unit testing
- val invoker = TransactionId(-100) // Invoker startup/shutdown or GC activity
- val invokerWarmup = TransactionId(-101) // Invoker warmup thread that makes
stem-cell containers
- val invokerNanny = TransactionId(-102) // Invoker nanny thread
- val dispatcher = TransactionId(-110) // Kafka message dispatcher
- val loadbalancer = TransactionId(-120) // Loadbalancer thread
- val invokerHealth = TransactionId(-121) // Invoker supervision
- val controller = TransactionId(-130) // Controller startup
- val dbBatcher = TransactionId(-140) // Database batcher
-
- def apply(tid: BigDecimal, extraLogging: Boolean = false): TransactionId = {
-Try {
- val now = Instant.now(Clock.systemUTC())
- TransactionId(TransactionMetadata(tid.toLong, now, extraLogging))
-} getOrElse unknown
+ val generatorConfig =
loadConfigOrThrow[TransactionGeneratorConfig](ConfigKeys.transactions)
+
+ val unknown = TransactionId("sid_unknown")
+ val testing = TransactionId("sid_testing") // Common id for for unit testing
+ val invoker = TransactionId("sid_invoker") // Invoker startup/shutdown or GC
activity
+ val invokerWarmup = TransactionId("sid_invokerWarmup") // Invoker warmup
thread that makes stem-cell containers
+ val invokerNanny = TransactionId("sid_invokerNanny") // Invoker nanny thread
+ val dispatcher = TransactionId("sid_dispatcher") // Kafka message dispatcher
+ val loadbalancer = TransactionId("sid_loadbalancer") // Loadbalancer thread
+ val invokerHealth = TransactionId("sid_invokerHealth") // Invoker supervision
+ val controller = TransactionId("sid_controller") // Controller startup
+ val dbBatcher = TransactionId("sid_dbBatcher") // Database batcher
+
+ def apply(tid: String, extraLogging: Boolean = false): TransactionId = {
+val now = Instant.now(Clock.systemUTC())
+TransactionId(TransactionMetadata(tid, now, extraLogging))
}
implicit val serdes = new RootJsonFormat[TransactionId] {
def write(t: TransactionId) = {
if (t.meta.extraLogging)
-JsArray(JsNumber(t.meta.id), JsNumber(t.meta.start.toEpochMilli),
JsBoolean(t.meta.extraLogging))
+JsArray(JsString(t.meta.id), JsNumber(t.meta.start.toEpochMilli),
JsBoolean(t.meta.extraLogging))
else
-JsArray(JsNumber(t.meta.id), JsNumber(t.meta.start.toEpochMilli))
+JsArray(JsString(t.meta.id), JsNumber(t.meta.start.toEpochMilli))
}
def read(value: JsValue) =
Try {
value match {
- case JsArray(Vector(JsNumber(id), JsNumber(start))) =>
-TransactionId(TransactionMetadata(id.longValue,
Instant.ofEpochMilli(start.longValue), false))
- case JsArray(Vector(JsNumber(id), JsNumber(start),
JsBoolean(extraLogging))) =>
-TransactionId(TransactionMetadata(id.longValue,
Instant.ofEpochMilli(start.longValue), extraLogging))
+ case JsArray(Vector(JsString(id), JsNumber(start))) =>
+TransactionId(TransactionMetadata(id,
Instant.ofEpochMilli(start.longValue), false))
+ case JsArray(Vector(JsString(id), JsNumber(start),
JsBoolean(extraLogging))) =>
+TransactionId(TransactionMetadata(id,
Instant.ofEpochMilli(start.longValue), extraLogging))
}
} getOrElse unknown
}
}
-/**
- * A thread-safe transaction counter.
- */
-trait TransactionCounter {
- case class TransactionCounterConfig(stride: Int)
-
- val transCounterConfig =
loadConfigOrThrow[TransactionCounterConfig](ConfigKeys.transactions)
- val stride = transCounterConfig.stride
- val instanceOrdinal: Int
-
- // seed the counter so transids do not overlap: instanceOrdinal + n *
stride, start at n = 1
- private lazy val cnt = new AtomicInteger(instanceOrdinal + stride)
-
- def transid(extraLogging: Boolean = false): TransactionId = {
-TransactionId(cnt.addAndGet(stride), extraLogging)
- }
-}
+case class TransactionGeneratorConfig(header: String)
Review comment:
I've addad a counter to the test. This might be helpful.
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:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] cbickel commented on a change in pull request #3199: One tid for the whole system.
cbickel commented on a change in pull request #3199: One tid for the whole
system.
URL:
https://github.com/apache/incubator-openwhisk/pull/3199#discussion_r182975010
##
File path: common/scala/src/main/scala/whisk/common/TransactionId.scala
##
@@ -200,58 +196,42 @@ object TransactionId {
val metricsKamonTags: Boolean =
sys.env.get("METRICS_KAMON_TAGS").getOrElse("False").toBoolean
val metricsLog: Boolean =
sys.env.get("METRICS_LOG").getOrElse("True").toBoolean
- val unknown = TransactionId(0)
- val testing = TransactionId(-1) // Common id for for unit testing
- val invoker = TransactionId(-100) // Invoker startup/shutdown or GC activity
- val invokerWarmup = TransactionId(-101) // Invoker warmup thread that makes
stem-cell containers
- val invokerNanny = TransactionId(-102) // Invoker nanny thread
- val dispatcher = TransactionId(-110) // Kafka message dispatcher
- val loadbalancer = TransactionId(-120) // Loadbalancer thread
- val invokerHealth = TransactionId(-121) // Invoker supervision
- val controller = TransactionId(-130) // Controller startup
- val dbBatcher = TransactionId(-140) // Database batcher
-
- def apply(tid: BigDecimal, extraLogging: Boolean = false): TransactionId = {
-Try {
- val now = Instant.now(Clock.systemUTC())
- TransactionId(TransactionMetadata(tid.toLong, now, extraLogging))
-} getOrElse unknown
+ val generatorConfig =
loadConfigOrThrow[TransactionGeneratorConfig](ConfigKeys.transactions)
+
+ val unknown = TransactionId("sid_unknown")
Review comment:
Good catch.
I've added a gate to the nginx and the controller.
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:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] cbickel commented on a change in pull request #3199: One tid for the whole system.
cbickel commented on a change in pull request #3199: One tid for the whole
system.
URL:
https://github.com/apache/incubator-openwhisk/pull/3199#discussion_r182972953
##
File path: ansible/group_vars/all
##
@@ -96,8 +96,10 @@ jmx:
jvmCommonArgs: "-Dcom.sun.management.jmxremote
-Dcom.sun.management.jmxremote.ssl=false
-Dcom.sun.management.jmxremote.local.only=false
-Dcom.sun.management.jmxremote.authenticate=true
-Dcom.sun.management.jmxremote.password.file=/root/jmxremote.password
-Dcom.sun.management.jmxremote.access.file=/root/jmxremote.access"
enabled: "{{ jmxremote_enabled | default('true') }}"
-transactions:
- stride: "{{ groups['controllers'] | length }}"
+transaction:
+ header: "{{ transaction_header | default('X-Request-ID') }}"
+ # The tid can be any String between 1 and 64 characters.
+ regex: "{{ whisk_tid_regex | default('[a-zA-Z0-9_-]{1,64}') }}"
Review comment:
That's a good idea. This makes operations much easier :)
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:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] markusthoemmes commented on issue #3416: Unit test PoolingRestClient
markusthoemmes commented on issue #3416: Unit test PoolingRestClient URL: https://github.com/apache/incubator-openwhisk/pull/3416#issuecomment-382996711 @dubee Have you thought about https://github.com/apache/incubator-openwhisk/pull/3416#discussion_r173771805? 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] markusthoemmes commented on issue #3504: use alpine-java base image (alpine 3.7 + oracle 8u162b12_server-jre)
markusthoemmes commented on issue #3504: use alpine-java base image (alpine 3.7 + oracle 8u162b12_server-jre) URL: https://github.com/apache/incubator-openwhisk/pull/3504#issuecomment-382989892 @tysonnorris Not sure on the image, there is an official openjdk image which has alpine tags? https://hub.docker.com/_/openjdk/ Any issues there? 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: us...@infra.apache.org With regards, Apache Git Services
[GitHub] chetanmeh opened a new pull request #3562: ArtifactStore implementation for CosmosDB
chetanmeh opened a new pull request #3562: ArtifactStore implementation for
CosmosDB
URL: https://github.com/apache/incubator-openwhisk/pull/3562
This PR provides a `CosmosDBArtifactStore` implementation to enable using
CosmosDB for storing subjects, whisks and activation in [Azure CosmosDB][1]
This is currently work in progress and opening this early so as to get
feedback on design progress
## Description
Currently OpenWhisk supports CouchDB for storing various entities like
actions, rules, subjects, activations etc. This PR provides a [CosmosDB][1]
based implementation for ArtifactStore SPI
### Usage of CosmosDB Java SDK
CosmosDB supports various modes to connect to it. For our usage we have two
options
1. [REST API][2]
2. [Async Java SDK][3]
3. [Sync Java SDK][6]
Compared to CouchDB performing queries against CosmosDB requires client side
computation which involves sending queries to each partition and then collect
and merge the result set. The Async Java SDK takes care of all these
interactions and provides a simplified reactive api based on [RxJava][5]
Given the complexity involved in performing various operations against
CosmosDB this PR uses the Java SDK to simplify and speed up the implementation
### Design Considerations
Partitioning Strategy
TBD
Query Execution
TBD
Attachment Storage
TBD
Testing Approach
TBD
## Related issue and scope
## My changes affect the following components
- [x] Data stores (e.g., CouchDB)
- [X] Tests
- [x] Deployment
- [X] Documentation
## Types of changes
- [x] Enhancement or new feature (adds new functionality).
## Checklist:
- [x] I signed an [Apache
CLA](https://github.com/apache/incubator-openwhisk/blob/master/CONTRIBUTING.md).
- [x] I reviewed the [style
guides](https://github.com/apache/incubator-openwhisk/wiki/Contributing:-Git-guidelines#code-readiness)
and followed the recommendations (Travis CI will check :).
- [x] I added tests to cover my changes.
- [ ] My changes require further changes to the documentation.
- [ ] I updated the documentation where necessary.
[1]: https://docs.microsoft.com/en-us/azure/cosmos-db/
[2]:
https://docs.microsoft.com/en-us/rest/api/documentdb/restful-interactions-with-documentdb-resources
[3]: https://docs.microsoft.com/en-us/azure/cosmos-db/sql-api-sdk-async-java
[4]: https://github.com/Azure/azure-cosmosdb-java
[5]: https://github.com/ReactiveX/RxJava
[6]: https://github.com/Azure/azure-documentdb-java
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:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] markusthoemmes commented on a change in pull request #3559: add support for mesos attribute constraints
markusthoemmes commented on a change in pull request #3559: add support for
mesos attribute constraints
URL:
https://github.com/apache/incubator-openwhisk/pull/3559#discussion_r182952956
##
File path:
common/scala/src/main/scala/whisk/core/mesos/MesosContainerFactory.scala
##
@@ -126,9 +139,28 @@ class MesosContainerFactory(config: WhiskConfig,
//strip any "--" prefixes on parameters (should make this consistent
everywhere else)
parameters
.map({ case (k, v) => if (k.startsWith("--")) (k.replaceFirst("--",
""), v) else (k, v) })
-++ containerArgs.extraArgs)
+++ containerArgs.extraArgs,
+ parseConstraints(constraintStrings))
}
+ /**
+ * Validate that constraint strings are well formed, and ignore constraints
with unknown operators
+ * @param constraintStrings
+ * @param logging
+ * @return
+ */
+ def parseConstraints(constraintStrings: Seq[String])(implicit logging:
Logging): Seq[Constraint] =
+constraintStrings.flatMap(cs => {
+ val parts = cs.split(mesosConfig.constraintDelimiter)
+ require(parts.length == 3, "constraint must be in the form
")
+ Seq(LIKE, UNLIKE).find(_.toString == parts(1)) match {
+case Some(o) => Some(Constraint(parts(0), o, parts(2)))
+case _ =>
+ logging.warn(this, s"ignoring unsupported constraint operator
${parts(1)}")
+ None
+ }
+})
Review comment:
Just to leave the note, this could make use of Scala's pattern matching, as
in:
```scala
cs.split(mesosConfig.constraintDelimiter).toList match {
case a :: LIKE.toString :: b => Some(Constraint(a, LIKE, b)
case a :: UNLIKE.toString :: b => Some(Constraint(a, UNLIKE, b))
case a ::o :: b=>
logging.warn(this, s"ignoring unsupported constraint operator $o")
None
case failure =>
logging.warn(this, s"couldn't parse constraint $failure")
None
}
```
This wouldn't throw an exception in any case and is typechecked throughout
(`parts(i)` will throw on an out-of-bounds exception)
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:
us...@infra.apache.org
With regards,
Apache Git Services
