aokolnychyi commented on code in PR #53143:
URL: https://github.com/apache/spark/pull/53143#discussion_r2573083077
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/CacheManager.scala:
##########
@@ -374,25 +376,68 @@ class CacheManager extends Logging with
AdaptiveSparkPlanHelper {
}
needToRecache.foreach { cd =>
cd.cachedRepresentation.cacheBuilder.clearCache()
- val sessionWithConfigsOff = getOrCloneSessionWithConfigsOff(spark)
- val (newKey, newCache) = sessionWithConfigsOff.withActive {
- val refreshedPlan = V2TableRefreshUtil.refresh(sessionWithConfigsOff,
cd.plan)
- val qe = sessionWithConfigsOff.sessionState.executePlan(refreshedPlan)
- qe.normalized ->
InMemoryRelation(cd.cachedRepresentation.cacheBuilder, qe)
- }
- val recomputedPlan = cd.copy(plan = newKey, cachedRepresentation =
newCache)
- this.synchronized {
- if (lookupCachedDataInternal(recomputedPlan.plan).nonEmpty) {
- logWarning("While recaching, data was already added to cache.")
- } else {
- cachedData = recomputedPlan +: cachedData
- CacheManager.logCacheOperation(log"Re-cached Dataframe cache entry:"
+
- log"${MDC(DATAFRAME_CACHE_ENTRY, recomputedPlan)}")
+ tryRebuildCacheEntry(spark, cd).foreach { entry =>
+ this.synchronized {
+ if (lookupCachedDataInternal(entry.plan).nonEmpty) {
+ logWarning("While recaching, data was already added to cache.")
+ } else {
+ cachedData = entry +: cachedData
+ CacheManager.logCacheOperation(log"Re-cached Dataframe cache
entry:" +
+ log"${MDC(DATAFRAME_CACHE_ENTRY, entry)}")
+ }
}
}
}
}
+ private def tryRebuildCacheEntry(spark: SparkSession, cd: CachedData):
Option[CachedData] = {
+ val sessionWithConfigsOff = getOrCloneSessionWithConfigsOff(spark)
+ sessionWithConfigsOff.withActive {
+ tryRefreshPlan(sessionWithConfigsOff, cd.plan).map { refreshedPlan =>
+ val qe = QueryExecution.create(
Review Comment:
This `QueryExecution` should not perform the refresh, meaning it won't
generate refresh-related exceptions. It will continue to throw other analysis
exceptions if any, as before.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
