easyice commented on code in PR #13828:
URL: https://github.com/apache/lucene/pull/13828#discussion_r1778298810
##
lucene/core/src/java/org/apache/lucene/util/bkd/DocIdsWriter.java:
##
@@ -205,12 +208,16 @@ void readInts(IndexInput in, int count, int[] docIDs)
throws IOExceptio
easyice commented on code in PR #13828:
URL: https://github.com/apache/lucene/pull/13828#discussion_r1778298810
##
lucene/core/src/java/org/apache/lucene/util/bkd/DocIdsWriter.java:
##
@@ -205,12 +208,16 @@ void readInts(IndexInput in, int count, int[] docIDs)
throws IOExceptio
easyice commented on code in PR #13828:
URL: https://github.com/apache/lucene/pull/13828#discussion_r1778291944
##
lucene/core/src/java/org/apache/lucene/util/bkd/DocIdsWriter.java:
##
@@ -205,12 +208,16 @@ void readInts(IndexInput in, int count, int[] docIDs)
throws IOExceptio
gf2121 commented on code in PR #13828:
URL: https://github.com/apache/lucene/pull/13828#discussion_r1778204945
##
lucene/core/src/java/org/apache/lucene/util/bkd/DocIdsWriter.java:
##
@@ -36,6 +38,7 @@ final class DocIdsWriter {
private static final byte LEGACY_DELTA_VINT = (
easyice opened a new pull request, #13828:
URL: https://github.com/apache/lucene/pull/13828
I got a memory allocation flamegraph from my cluster, It shows the
`DocIdsWriter#readBitSetIterator` allocating about ~14GB RAM in 30 seconds,
this proposal uses a `scratchLongs[]` array instead of
