Github user andrewor14 commented on a diff in the pull request:
https://github.com/apache/spark/pull/8639#discussion_r42300566
--- Diff:
core/src/main/scala/org/apache/spark/scheduler/cluster/mesos/MesosSchedulerBackend.scala
---
@@ -207,23 +211,44 @@ private[spark] class MesosSchedulerBackend(
*/
override def resourceOffers(d: SchedulerDriver, offers: JList[Offer]) {
inClassLoader() {
+ // Fail first on offers with unmet constraints
+ val (offersMatchingConstraints, offersNotMatchingConstraints) =
+ offers.asScala.partition { o =>
+ val offerAttributes = toAttributeMap(o.getAttributesList)
+ val meetsConstraints =
+ matchesAttributeRequirements(slaveOfferConstraints,
offerAttributes)
+
+ // add some debug messaging
+ if (!meetsConstraints) {
+ val id = o.getId.getValue
+ logDebug(s"Declining offer: $id with attributes:
$offerAttributes")
+ }
+
+ meetsConstraints
+ }
+
+ // These offers do not meet constraints. We don't need to see them
again.
+ // Decline the offer for a long period of time.
+ offersNotMatchingConstraints.foreach { o =>
+ d.declineOffer(o.getId, Filters.newBuilder()
+
.setRefuseSeconds(rejectOfferDurationForUnmetConstraints).build())
+ }
+
// Fail-fast on offers we know will be rejected
--- End diff --
this comment is outdated:
```
// Of the matching constraints, see which ones give us enough memory and
cores
```
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
