HomeMesosphereNo notifications. 4 unresolved issues.
  • Search
Differential D297

fixes #4808, relocate pod unreachbleStrategy to PodSchedulingPolicy type
ClosedAll Users
Actions

Authored by jdef on Dec 8 2016, 3:38 PM.

Details

Reviewers
aquamatthias
jeschkies
Commits
rMARATHON38a0481b3842: fixes #4808, relocate pod unreachbleStrategy to PodSchedulingPolicy type
rMARATHON4c83608a27c6: fixes #4808, relocate pod unreachbleStrategy to PodSchedulingPolicy type
rMARATHONa5cdd6fe6aa8: fixes #4808, relocate pod unreachbleStrategy to PodSchedulingPolicy type
Summary

https://github.com/jdef

Test Plan

sbt test

Diff Detail

Repository
rMARATHON marathon
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.
Changes from before your most recent comment are hidden. Show Older Changes
jeschkies accepted this revision.Dec 8 2016, 4:09 PM

Thanks for moving this. Could you create a Github issue instead of the link?

src/test/scala/mesosphere/marathon/integration/TaskUnreachableIntegrationTest.scala
30–31

Is there a Github issue for this todo?

This revision is now accepted and ready to land.Dec 8 2016, 4:09 PM
aquamatthias accepted this revision.Dec 8 2016, 4:26 PM
Closed by commit rMARATHON4c83608a27c6: fixes #4808, relocate pod unreachbleStrategy to PodSchedulingPolicy type (authored by jdef, committed by aquamatthias). · Explain WhyDec 8 2016, 4:43 PM
This revision was automatically updated to reflect the committed changes.

Revision Contents

Diff 1159

View Options

docs/docs/rest-api/public/api/v2/types/pod.raml

Show First 20 Lines • Show All 106 Lines • ▼ Show 20 Line(s)
107107 See [the Mesos attribute and resources documentation](http://mesos.apache.org/documentation/latest/attributes-resources/) for details
108108 PodSchedulingPolicy:
109109 type: object
110110 properties:
111111 backoff?: PodSchedulingBackoffStrategy
112112 upgrade?: PodUpgradeStrategy
113113 placement?: PodPlacementPolicy
114114 killSelection?: killSelection.KillSelection
115 unreachableStrategy?: unreachableStrategy.UnreachableStrategy
115116 Pod:
116117 description: |
117118 A pod allows one to launch a collection co-located (on the same agent) containers
118119 that share the same network namespace and that may share the same volumes.
119120 Resources are specified on a per-container basis.
120121 properties:
121122 id: strings.PathId
122123 labels?:
Show All 21 Lines
144145 description: Volumes defined on a pod level that are mounted into containers
145146 networks?:
146147 type: net.Network[]
147148 description: |
148149 Network settings are defined on a pod level. All containers share the same network stack.
149150 At this time, only one stack is supported.
150151 scaling?: PodScalingPolicy
151152 scheduling?: PodSchedulingPolicy
152 unreachableStrategy?: unreachableStrategy.UnreachableStrategy
153153 executorResources?:
154154 type: resources.ExecutorResources
155155 description: The resources to allocate to the executor.
View Options

src/main/scala/mesosphere/marathon/raml/PodConversion.scala

Show All 19 Lines
2020
2121 val resourceRoles = podDef.scheduling.flatMap(_.placement)
2222 .fold(Set.empty[String])(_.acceptedResourceRoles.toSet)
2323
2424 val upgradeStrategy = podDef.scheduling.flatMap(_.upgrade).fold(DefaultUpgradeStrategy) { raml =>
2525 state.UpgradeStrategy(raml.minimumHealthCapacity, raml.maximumOverCapacity)
2626 }
2727
28 val unreachableStrategy = podDef.unreachableStrategy.fold(DefaultUnreachableStrategy)(Raml.fromRaml(_))
28 val unreachableStrategy = podDef.scheduling.flatMap(_.unreachableStrategy).fold(DefaultUnreachableStrategy)(Raml.fromRaml(_))
2929 val killSelection: state.KillSelection = podDef.scheduling.fold(state.KillSelection.DefaultKillSelection) {
3030 _.killSelection.fold(state.KillSelection.DefaultKillSelection)(Raml.fromRaml(_))
3131 }
3232
3333 val backoffStrategy = podDef.scheduling.flatMap { policy =>
3434 policy.backoff.map { strategy =>
3535 state.BackoffStrategy(strategy.backoff.seconds, strategy.maxLaunchDelay.seconds, strategy.backoffFactor)
3636 }
Show All 32 Lines
6969 val ramlUpgradeStrategy = PodUpgradeStrategy(
7070 pod.upgradeStrategy.minimumHealthCapacity,
7171 pod.upgradeStrategy.maximumOverCapacity)
7272
7373 val ramlBackoffStrategy = PodSchedulingBackoffStrategy(
7474 backoff = pod.backoffStrategy.backoff.toMillis.toDouble / 1000.0,
7575 maxLaunchDelay = pod.backoffStrategy.maxLaunchDelay.toMillis.toDouble / 1000.0,
7676 backoffFactor = pod.backoffStrategy.factor)
77 val schedulingPolicy = PodSchedulingPolicy(Some(ramlBackoffStrategy), Some(ramlUpgradeStrategy),
77 val schedulingPolicy = PodSchedulingPolicy(
78 Some(ramlBackoffStrategy),
79 Some(ramlUpgradeStrategy),
7880 Some(PodPlacementPolicy(
7981 pod.constraints.toRaml[Seq[Constraint]],
80 pod.acceptedResourceRoles.toIndexedSeq)))
82 pod.acceptedResourceRoles.toIndexedSeq)),
83 Some(pod.killSelection.toRaml),
84 Some(pod.unreachableStrategy.toRaml)
85 )
8186
8287 val scalingPolicy = FixedPodScalingPolicy(pod.instances, pod.maxInstances)
8388
8489 Pod(
8590 id = pod.id.toString,
8691 version = Some(pod.version.toOffsetDateTime),
8792 user = pod.user,
8893 containers = pod.containers.map(Raml.toRaml(_)),
Show All 29 Lines
View Options

src/test/scala/mesosphere/marathon/integration/TaskUnreachableIntegrationTest.scala

Show All 21 Lines
2222 "scale_apps_initial_delay" -> "5000",
2323 "scale_apps_interval" -> "5000",
2424 "min_revive_offers_interval" -> "100",
2525 "task_lost_expunge_gc" -> "30000",
2626 "task_lost_expunge_initial_delay" -> "1000",
2727 "task_lost_expunge_interval" -> "1000"
2828 )
2929
30 // TODO unreachable tests for pods
31
jeschkiesUnsubmitted
Not Done

Is there a Github issue for this todo?

3032 before {
3133 mesosCluster.agents(1).stop()
3234 mesosCluster.masters(1).stop()
3335 mesosCluster.masters.head.start()
3436 mesosCluster.agents.head.start()
3537 mesosCluster.waitForLeader().futureValue
3638 }
3739
▲ Show 20 Lines • Show All 155 Lines • Show Last 20 Lines