Github user nishkamravi2 commented on a diff in the pull request:
https://github.com/apache/spark/pull/7274#discussion_r34981074
--- Diff: core/src/main/scala/org/apache/spark/deploy/master/Master.scala
---
@@ -543,39 +544,72 @@ private[master] class Master(
* multiple executors from the same application may be launched on the
same worker if the worker
* has enough cores and memory. Otherwise, each executor grabs all the
cores available on the
* worker by default, in which case only one executor may be launched on
each worker.
+ *
+ * It is important to allocate coresPerExecutor on each worker at a time
(instead of 1 core
+ * at a time). Consider the following example: cluster has 4 workers
with 16 cores each.
+ * User requests 3 executors (spark.cores.max = 48, spark.executor.cores
= 16). If 1 core is
+ * allocated at a time, 12 cores from each worker would be assigned to
each executor.
+ * Since 12 < 16, no executors would launch [SPARK-8881].
*/
- private def startExecutorsOnWorkers(): Unit = {
- // Right now this is a very simple FIFO scheduler. We keep trying to
fit in the first app
- // in the queue, then the second app, etc.
+ private[master] def scheduleExecutorsOnWorkers(
+ app: ApplicationInfo,
+ usableWorkers: Array[WorkerInfo],
+ spreadOutApps: Boolean): Array[Int] = {
+ // If the number of cores per executor is not specified, then we can
just schedule
+ // 1 core at a time since we expect a single executor to be launched
on each worker
+ val coresPerExecutor = app.desc.coresPerExecutor.getOrElse(1)
+ val memoryPerExecutor = app.desc.memoryPerExecutorMB
+ val numUsable = usableWorkers.length
+ val assignedCores = new Array[Int](numUsable) // Number of cores to
give to each worker
+ val assignedMemory = new Array[Int](numUsable) // Amount of memory to
give to each worker
+ var coresToAssign = math.min(app.coresLeft,
usableWorkers.map(_.coresFree).sum)
+ var pos = 0
if (spreadOutApps) {
- // Try to spread out each app among all the workers, until it has
all its cores
- for (app <- waitingApps if app.coresLeft > 0) {
- val usableWorkers = workers.toArray.filter(_.state ==
WorkerState.ALIVE)
- .filter(worker => worker.memoryFree >=
app.desc.memoryPerExecutorMB &&
- worker.coresFree >= app.desc.coresPerExecutor.getOrElse(1))
- .sortBy(_.coresFree).reverse
- val numUsable = usableWorkers.length
- val assigned = new Array[Int](numUsable) // Number of cores to
give on each node
- var toAssign = math.min(app.coresLeft,
usableWorkers.map(_.coresFree).sum)
- var pos = 0
- while (toAssign > 0) {
- if (usableWorkers(pos).coresFree - assigned(pos) > 0) {
- toAssign -= 1
- assigned(pos) += 1
- }
- pos = (pos + 1) % numUsable
- }
- // Now that we've decided how many cores to give on each node,
let's actually give them
- for (pos <- 0 until numUsable if assigned(pos) > 0) {
- allocateWorkerResourceToExecutors(app, assigned(pos),
usableWorkers(pos))
+ // Try to spread out executors among workers (sparse scheduling)
+ while (coresToAssign > 0) {
+ if (usableWorkers(pos).coresFree - assignedCores(pos) >=
coresPerExecutor &&
+ usableWorkers(pos).memoryFree - assignedMemory(pos) >=
memoryPerExecutor) {
+ coresToAssign -= coresPerExecutor
+ assignedCores(pos) += coresPerExecutor
+ assignedMemory(pos) += memoryPerExecutor
--- End diff --
"resources that looked available aren't anymore and fail to schedule,
that's fine." This is the assumption being made here. If the user didn't care
about the size of the executor, they would skip executor.cores and the
algorithm would proceed as before (best-effort: one-core at a time). If they
do, we should either schedule as requested or not at all. If we care to be
extra-friendly, we could add a check to log a message from within the loop:
"Not enough resources, please check spark.cores.max and spark.executor.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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
