cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From alek...@apache.org
Subject [2/2] git commit: Merge branch 'cassandra-1.2' into cassandra-2.0
Date Tue, 10 Sep 2013 14:58:37 GMT
Merge branch 'cassandra-1.2' into cassandra-2.0

Conflicts:
	CHANGES.txt
	src/java/org/apache/cassandra/db/index/composites/CompositesSearcher.java


Project: http://git-wip-us.apache.org/repos/asf/cassandra/repo
Commit: http://git-wip-us.apache.org/repos/asf/cassandra/commit/096f2bf9
Tree: http://git-wip-us.apache.org/repos/asf/cassandra/tree/096f2bf9
Diff: http://git-wip-us.apache.org/repos/asf/cassandra/diff/096f2bf9

Branch: refs/heads/cassandra-2.0
Commit: 096f2bf92049ad3d955eb1189526800c8292dd94
Parents: 52678a2 caef32e
Author: Aleksey Yeschenko <aleksey@apache.org>
Authored: Tue Sep 10 17:58:19 2013 +0300
Committer: Aleksey Yeschenko <aleksey@apache.org>
Committed: Tue Sep 10 17:58:19 2013 +0300

----------------------------------------------------------------------
 CHANGES.txt                                     |  2 ++
 .../apache/cassandra/db/ColumnFamilyStore.java  |  2 +-
 .../db/index/composites/CompositesSearcher.java | 26 +++++++++-----------
 3 files changed, 14 insertions(+), 16 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/cassandra/blob/096f2bf9/CHANGES.txt
----------------------------------------------------------------------
diff --cc CHANGES.txt
index abbb4f9,2328bf7..22fa74b
--- a/CHANGES.txt
+++ b/CHANGES.txt
@@@ -23,54 -7,15 +23,56 @@@ Merged from 1.2
   * Allow disabling SlabAllocator (CASSANDRA-5935)
   * Make user-defined compaction JMX blocking (CASSANDRA-4952)
   * Fix streaming does not transfer wrapped range (CASSANDRA-5948)
+  * Fix loading index summary containing empty key (CASSANDRA-5965)
+  * Correctly handle limits in CompositesSearcher (CASSANDRA-5975)
  
  
 -1.2.9
 +2.0.0
 + * Fix thrift validation when inserting into CQL3 tables (CASSANDRA-5138)
 + * Fix periodic memtable flushing behavior with clean memtables (CASSANDRA-5931)
 + * Fix dateOf() function for pre-2.0 timestamp columns (CASSANDRA-5928)
 + * Fix SSTable unintentionally loads BF when opened for batch (CASSANDRA-5938)
 + * Add stream session progress to JMX (CASSANDRA-4757)
 + * Fix NPE during CAS operation (CASSANDRA-5925)
 +Merged from 1.2:
   * Fix getBloomFilterDiskSpaceUsed for AlwaysPresentFilter (CASSANDRA-5900)
 - * migrate 1.1 schema_columnfamilies.key_alias column to key_aliases
 -   (CASSANDRA-5800)
 - * add --migrate option to sstableupgrade and sstablescrub (CASSANDRA-5831)
 + * Don't announce schema version until we've loaded the changes locally
 +   (CASSANDRA-5904)
 + * Fix to support off heap bloom filters size greater than 2 GB (CASSANDRA-5903)
 + * Properly handle parsing huge map and set literals (CASSANDRA-5893)
 +
 +
 +2.0.0-rc2
 + * enable vnodes by default (CASSANDRA-5869)
 + * fix CAS contention timeout (CASSANDRA-5830)
 + * fix HsHa to respect max frame size (CASSANDRA-4573)
 + * Fix (some) 2i on composite components omissions (CASSANDRA-5851)
 + * cqlsh: add DESCRIBE FULL SCHEMA variant (CASSANDRA-5880)
 +Merged from 1.2:
 + * Correctly validate sparse composite cells in scrub (CASSANDRA-5855)
 + * Add KeyCacheHitRate metric to CF metrics (CASSANDRA-5868)
 + * cqlsh: add support for multiline comments (CASSANDRA-5798)
 + * Handle CQL3 SELECT duplicate IN restrictions on clustering columns
 +   (CASSANDRA-5856)
 +
 +
 +2.0.0-rc1
 + * improve DecimalSerializer performance (CASSANDRA-5837)
 + * fix potential spurious wakeup in AsyncOneResponse (CASSANDRA-5690)
 + * fix schema-related trigger issues (CASSANDRA-5774)
 + * Better validation when accessing CQL3 table from thrift (CASSANDRA-5138)
 + * Fix assertion error during repair (CASSANDRA-5801)
 + * Fix range tombstone bug (CASSANDRA-5805)
 + * DC-local CAS (CASSANDRA-5797)
 + * Add a native_protocol_version column to the system.local table (CASSANRDA-5819)
 + * Use index_interval from cassandra.yaml when upgraded (CASSANDRA-5822)
 + * Fix buffer underflow on socket close (CASSANDRA-5792)
 +Merged from 1.2:
 + * Fix reading DeletionTime from 1.1-format sstables (CASSANDRA-5814)
 + * cqlsh: add collections support to COPY (CASSANDRA-5698)
 + * retry important messages for any IOException (CASSANDRA-5804)
 + * Allow empty IN relations in SELECT/UPDATE/DELETE statements (CASSANDRA-5626)
 + * cqlsh: fix crashing on Windows due to libedit detection (CASSANDRA-5812)
   * fix bulk-loading compressed sstables (CASSANDRA-5820)
   * (Hadoop) fix quoting in CqlPagingRecordReader and CqlRecordWriter 
     (CASSANDRA-5824)

http://git-wip-us.apache.org/repos/asf/cassandra/blob/096f2bf9/src/java/org/apache/cassandra/db/ColumnFamilyStore.java
----------------------------------------------------------------------

http://git-wip-us.apache.org/repos/asf/cassandra/blob/096f2bf9/src/java/org/apache/cassandra/db/index/composites/CompositesSearcher.java
----------------------------------------------------------------------
diff --cc src/java/org/apache/cassandra/db/index/composites/CompositesSearcher.java
index f9b7b11,1e9d59d..011839e
--- a/src/java/org/apache/cassandra/db/index/composites/CompositesSearcher.java
+++ b/src/java/org/apache/cassandra/db/index/composites/CompositesSearcher.java
@@@ -101,12 -170,15 +101,14 @@@ public class CompositesSearcher extend
          return new ColumnFamilyStore.AbstractScanIterator()
          {
              private ByteBuffer lastSeenPrefix = startPrefix;
 -            private ArrayDeque<IColumn> indexColumns;
 -            private final QueryPath path = new QueryPath(baseCfs.columnFamily);
 +            private Deque<Column> indexColumns;
              private int columnsRead = Integer.MAX_VALUE;
 -            private int limit = ((SliceQueryFilter)filter.initialFilter()).count;
++            private int limit = filter.currentLimit();
+             private int columnsCount = 0;
  
-             private final int meanColumns = Math.max(index.getIndexCfs().getMeanColumns(),
1);
+             private int meanColumns = Math.max(index.getIndexCfs().getMeanColumns(), 1);
              // We shouldn't fetch only 1 row as this provides buggy paging in case the first
row doesn't satisfy all clauses
--            private final int rowsPerQuery = Math.max(Math.min(filter.maxRows(), filter.maxColumns()
/ meanColumns), 2);
++            private int rowsPerQuery = Math.max(Math.min(filter.maxRows(), filter.maxColumns()
/ meanColumns), 2);
  
              public boolean needsFiltering()
              {
@@@ -138,11 -206,9 +136,9 @@@
  
                  while (true)
                  {
-                     // Did we got more columns that needed to respect the user limit?
-                     // (but we still need to return what was fetch already)
+                     // Did we get more columns that needed to respect the user limit?
+                     // (but we still need to return what has been fetched already)
 -                    if (columnsCount >= limit)
 +                    if (columnsCount > limit)
                          return makeReturn(currentKey, data);
  
                      if (indexColumns == null || indexColumns.isEmpty())
@@@ -162,16 -228,15 +158,16 @@@
                                                                               lastSeenPrefix,
                                                                               endPrefix,
                                                                               false,
 -                                                                             rowsPerQuery);
 +                                                                             rowsPerQuery,
 +                                                                             filter.timestamp);
                          ColumnFamily indexRow = index.getIndexCfs().getColumnFamily(indexFilter);
 -                        if (indexRow == null)
 +                        if (indexRow == null || indexRow.getColumnCount() == 0)
                              return makeReturn(currentKey, data);
  
 -                        Collection<IColumn> sortedColumns = indexRow.getSortedColumns();
 +                        Collection<Column> sortedColumns = indexRow.getSortedColumns();
                          columnsRead = sortedColumns.size();
-                         indexColumns = new ArrayDeque<Column>(sortedColumns);
 -                        indexColumns = new ArrayDeque<IColumn>(sortedColumns);
 -                        IColumn firstColumn = sortedColumns.iterator().next();
++                        indexColumns = new ArrayDeque<>(sortedColumns);
 +                        Column firstColumn = sortedColumns.iterator().next();
  
                          // Paging is racy, so it is possible the first column of a page
is not the last seen one.
                          if (lastSeenPrefix != startPrefix && lastSeenPrefix.equals(firstColumn.name()))


Mime
View raw message