Implement sstable generation identifier as uuid

Patch by Jacek Lewandowski; reviewed by Andrés de la Peña, Benjamin Lerer and Dan Jatnieks for CASSANDRA-17048

.build/build-rat.xml

@@ -68,6 +68,7 @@
                  <exclude name="**/test/data/jmxdump/cassandra-3.0-jmx.yaml"/>
                  <exclude name="**/test/data/jmxdump/cassandra-3.11-jmx.yaml"/>
                  <exclude name="**/test/data/jmxdump/cassandra-4.0-jmx.yaml"/>
+                 <exclude name="**/test/data/jmxdump/cassandra-4.1-jmx.yaml"/>
                  <exclude name="**/test/resources/data/config/YamlConfigurationLoaderTest/shared_client_error_reporting_exclusions.yaml"/>
                  <exclude name="**/tools/cqlstress-counter-example.yaml"/>
                  <exclude name="**/tools/cqlstress-example.yaml"/>

CHANGES.txt

@@ -1,4 +1,5 @@
 4.1
+ * Add support for UUID based sstable generation identifiers (CASSANDRA-17048)
  * Log largest memtable flush at info instead of debug (CASSANDRA-17472)
  * Add native transport rate limiter options to example cassandra.yaml, and expose metric for dispatch rate (CASSANDRA-17423)
  * Add diagnostic events for guardrails (CASSANDRA-17197)

NEWS.txt

@@ -116,6 +116,12 @@ New features
       table.
     - Added ability to invalidate auth caches through corresponding `nodetool` commands and virtual tables.
     - DCL statements in audit logs will now obscure only the password if they don't fail to parse.
+    - Starting from 4.1 sstables support UUID based generation identifiers. They are globally unique and thus they let
+      the node to create sstables without any prior knowledge about the existing sstables in the data directory.
+      The feature is disabled by default in cassandra.yaml because once enabled, there is no easy way to downgrade.
+      When the node is restarted with UUID based generation identifiers enabled, each newly created sstable will have
+      a UUID based generation identifier and such files are not readable by previous Cassandra versions. In the future
+      those new identifiers will become enabled by default.
 
 Upgrading
 ---------

conf/cassandra.yaml

@@ -947,6 +947,13 @@ compaction_throughput: 64MiB/s
 # between the sstables, reducing page cache churn and keeping hot rows hot
 sstable_preemptive_open_interval: 50MiB
 
+# Starting from 4.1 sstables support UUID based generation identifiers. They are disabled by default
+# because once enabled, there is no easy way to downgrade. When the node is restarted with this option
+# set to true, each newly created sstable will have a UUID based generation identifier and such files are
+# not readable by previous Cassandra versions. At some point, this option will become true by default
+# and eventually get removed from the configuration.
+enable_uuid_sstable_identifiers: false
+
 # When enabled, permits Cassandra to zero-copy stream entire eligible
 # SSTables between nodes, including every component.
 # This speeds up the network transfer significantly subject to
@@ -1628,7 +1635,7 @@ drop_compact_storage_enabled: false
 # removing list elements by either index or value. Defaults to true.
 # read_before_write_list_operations_enabled: true
 # Guardrail to warn or fail when querying with an IN restriction selecting more partition keys than threshold.
-# The two thresholds default to -1 to disable. 
+# The two thresholds default to -1 to disable.
 # partition_keys_in_select_warn_threshold: -1
 # partition_keys_in_select_fail_threshold: -1
 # Guardrail to warn or fail when an IN query creates a cartesian product with a size exceeding threshold,
@@ -1672,3 +1679,4 @@ drop_compact_storage_enabled: false
 #    enabled: true # (overriden by cassandra.ignore_dc system property)
 #  rack:
 #    enabled: true # (overriden by cassandra.ignore_rack system property)
+

src/java/org/apache/cassandra/config/Config.java

@@ -748,6 +748,9 @@ public static boolean isClientMode()
     public volatile boolean auto_optimise_full_repair_streams = false;
     public volatile boolean auto_optimise_preview_repair_streams = false;
 
+    // see CASSANDRA-17048 and the comment in cassandra.yaml
+    public boolean enable_uuid_sstable_identifiers = false;
+
     /**
      * Client mode means that the process is a pure client, that uses C* code base but does
      * not read or write local C* database files.

src/java/org/apache/cassandra/config/DatabaseDescriptor.java

@@ -18,9 +18,22 @@
 package org.apache.cassandra.config;
 
 import java.io.IOException;
-import java.net.*;
+import java.net.Inet4Address;
+import java.net.Inet6Address;
+import java.net.InetAddress;
+import java.net.NetworkInterface;
+import java.net.SocketException;
+import java.net.UnknownHostException;
 import java.nio.file.FileStore;
-import java.util.*;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Comparator;
+import java.util.Enumeration;
+import java.util.List;
+import java.util.Map;
+import java.util.Objects;
+import java.util.Set;
+import java.util.UUID;
 import java.util.concurrent.TimeUnit;
 import java.util.function.Function;
 import java.util.function.Supplier;
@@ -31,18 +44,12 @@
 import com.google.common.primitives.Ints;
 import com.google.common.primitives.Longs;
 import com.google.common.util.concurrent.RateLimiter;
-
-import org.apache.cassandra.config.Config.PaxosOnLinearizabilityViolation;
-import org.apache.cassandra.db.ConsistencyLevel;
-import org.apache.cassandra.gms.IFailureDetector;
-import org.apache.cassandra.io.util.File;
-import org.apache.cassandra.config.Config.PaxosStatePurging;
-
+import org.apache.commons.lang3.ArrayUtils;
+import org.apache.commons.lang3.StringUtils;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 import org.apache.cassandra.audit.AuditLogOptions;
-import org.apache.cassandra.fql.FullQueryLoggerOptions;
 import org.apache.cassandra.auth.AllowAllInternodeAuthenticator;
 import org.apache.cassandra.auth.AuthConfig;
 import org.apache.cassandra.auth.IAuthenticator;
@@ -51,14 +58,20 @@
 import org.apache.cassandra.auth.INetworkAuthorizer;
 import org.apache.cassandra.auth.IRoleManager;
 import org.apache.cassandra.config.Config.CommitLogSync;
+import org.apache.cassandra.config.Config.PaxosOnLinearizabilityViolation;
+import org.apache.cassandra.config.Config.PaxosStatePurging;
+import org.apache.cassandra.db.ConsistencyLevel;
 import org.apache.cassandra.db.commitlog.AbstractCommitLogSegmentManager;
 import org.apache.cassandra.db.commitlog.CommitLog;
 import org.apache.cassandra.db.commitlog.CommitLogSegmentManagerCDC;
 import org.apache.cassandra.db.commitlog.CommitLogSegmentManagerStandard;
 import org.apache.cassandra.dht.IPartitioner;
 import org.apache.cassandra.exceptions.ConfigurationException;
+import org.apache.cassandra.fql.FullQueryLoggerOptions;
+import org.apache.cassandra.gms.IFailureDetector;
 import org.apache.cassandra.io.FSWriteError;
 import org.apache.cassandra.io.util.DiskOptimizationStrategy;
+import org.apache.cassandra.io.util.File;
 import org.apache.cassandra.io.util.FileUtils;
 import org.apache.cassandra.io.util.PathUtils;
 import org.apache.cassandra.io.util.SpinningDiskOptimizationStrategy;
@@ -72,22 +85,14 @@
 import org.apache.cassandra.security.EncryptionContext;
 import org.apache.cassandra.security.SSLFactory;
 import org.apache.cassandra.service.CacheService.CacheType;
-import org.apache.cassandra.service.FileSystemOwnershipCheck;
-import org.apache.cassandra.service.StartupChecks;
 import org.apache.cassandra.service.paxos.Paxos;
 import org.apache.cassandra.utils.FBUtilities;
 
-import org.apache.commons.lang3.ArrayUtils;
-import org.apache.commons.lang3.StringUtils;
-
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
-
 import static org.apache.cassandra.config.CassandraRelevantProperties.OS_ARCH;
 import static org.apache.cassandra.config.CassandraRelevantProperties.SUN_ARCH_DATA_MODEL;
+import static org.apache.cassandra.config.CassandraRelevantProperties.TEST_JVM_DTEST_DISABLE_SSL;
 import static org.apache.cassandra.io.util.FileUtils.ONE_GIB;
 import static org.apache.cassandra.io.util.FileUtils.ONE_MIB;
-import static org.apache.cassandra.config.CassandraRelevantProperties.TEST_JVM_DTEST_DISABLE_SSL;
 import static org.apache.cassandra.utils.Clock.Global.logInitializationOutcome;
 
 public class DatabaseDescriptor
@@ -4086,4 +4091,9 @@ public static void setStreamingStateSize(DataStorageSpec duration)
             conf.streaming_state_size = duration;
         }
     }
+
+    public static boolean isUUIDSSTableIdentifiersEnabled()
+    {
+        return conf.enable_uuid_sstable_identifiers;
+    }
 }