This is an automated email from the ASF dual-hosted git repository.
jmckenzie pushed a commit to branch trunk
in repository https://gitbox.apache.org/repos/asf/cassandra.git
The following commit(s) were added to refs/heads/trunk by this push:
new f4a2135 Remove strong ref loop in LogTransaction.SSTableTidier
f4a2135 is described below
commit f4a2135c5ba442aafd27bb7c12c85b376d5a2b87
Author: Josh McKenzie <[email protected]>
AuthorDate: Tue Dec 14 12:01:47 2021 -0500
Remove strong ref loop in LogTransaction.SSTableTidier
Patch by Josh McKenzie; reviewed by Benedict Elliott Smith for
CASSANDRA-17205
---
CHANGES.txt | 9 ++++---
.../cassandra/db/lifecycle/LogTransaction.java | 29 +++++++++++++++-------
2 files changed, 25 insertions(+), 13 deletions(-)
diff --git a/CHANGES.txt b/CHANGES.txt
index 5a99669..e829293 100644
--- a/CHANGES.txt
+++ b/CHANGES.txt
@@ -1,7 +1,8 @@
4.1
+ * Remove self-reference in SSTableTidier (CASSANDRA-17205)
* Add guardrail for query page size (CASSANDRA-17189)
* Allow column_index_size_in_kb to be configurable through nodetool
(CASSANDRA-17121)
- * Emit a metric for number of local read and write calls
+ * Emit a metric for number of local read and write calls
* Add non-blocking mode for CDC writes (CASSANDRA-17001)
* Add guardrails framework (CASSANDRA-17147)
* Harden resource management on SSTable components to prevent future leaks
(CASSANDRA-17174)
@@ -228,7 +229,7 @@ Merged from 3.0:
* Promote protocol V5 out of beta (CASSANDRA-14973)
* Fix incorrect encoding for strings can be UTF8 (CASSANDRA-16429)
* Fix node unable to join when RF > N in multi-DC with added warning
(CASSANDRA-16296)
- * Add an option to nodetool tablestats to check sstable location correctness
(CASSANDRA-16344)
+ * Add an option to nodetool tablestats to check sstable location correctness
(CASSANDRA-16344)
* Unable to ALTER KEYSPACE while decommissioned/assassinated nodes are in
gossip (CASSANDRA-16422)
* Metrics backward compatibility restored after CASSANDRA-15066
(CASSANDRA-16083)
* Reduce new reserved keywords introduced since 3.0 (CASSANDRA-16439)
@@ -768,7 +769,7 @@ Merged from 3.0:
* Use standard Amazon naming for datacenter and rack in Ec2Snitch
(CASSANDRA-7839)
* Abstract write path for pluggable storage (CASSANDRA-14118)
* nodetool describecluster should be more informative (CASSANDRA-13853)
- * Compaction performance improvements (CASSANDRA-14261)
+ * Compaction performance improvements (CASSANDRA-14261)
* Refactor Pair usage to avoid boxing ints/longs (CASSANDRA-14260)
* Add options to nodetool tablestats to sort and limit output
(CASSANDRA-13889)
* Rename internals to reflect CQL vocabulary (CASSANDRA-14354)
@@ -1153,7 +1154,7 @@ Merged from 2.1:
3.11.2
* Fix ReadCommandTest (CASSANDRA-14234)
* Remove trailing period from latency reports at keyspace level
(CASSANDRA-14233)
- * Remove dependencies on JVM internal classes from JMXServerUtils
(CASSANDRA-14173)
+ * Remove dependencies on JVM internal classes from JMXServerUtils
(CASSANDRA-14173)
* Add DEFAULT, UNSET, MBEAN and MBEANS to `ReservedKeywords` (CASSANDRA-14205)
* Print correct snitch info from nodetool describecluster (CASSANDRA-13528)
* Enable CDC unittest (CASSANDRA-14141)
diff --git a/src/java/org/apache/cassandra/db/lifecycle/LogTransaction.java
b/src/java/org/apache/cassandra/db/lifecycle/LogTransaction.java
index 09717fc..5a909e0 100644
--- a/src/java/org/apache/cassandra/db/lifecycle/LogTransaction.java
+++ b/src/java/org/apache/cassandra/db/lifecycle/LogTransaction.java
@@ -30,12 +30,12 @@ import java.util.function.Predicate;
import com.google.common.annotations.VisibleForTesting;
import com.google.common.util.concurrent.Runnables;
-import org.apache.cassandra.io.util.File;
+import com.codahale.metrics.Counter;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.apache.cassandra.concurrent.ScheduledExecutors;
-import org.apache.cassandra.schema.TableMetadata;
+import org.apache.cassandra.config.DatabaseDescriptor;
import org.apache.cassandra.db.Directories;
import org.apache.cassandra.db.SystemKeyspace;
import org.apache.cassandra.db.compaction.OperationType;
@@ -46,7 +46,9 @@ import org.apache.cassandra.io.sstable.Descriptor;
import org.apache.cassandra.io.sstable.SSTable;
import org.apache.cassandra.io.sstable.SnapshotDeletingTask;
import org.apache.cassandra.io.sstable.format.SSTableReader;
+import org.apache.cassandra.io.util.File;
import org.apache.cassandra.io.util.FileUtils;
+import org.apache.cassandra.schema.TableMetadata;
import org.apache.cassandra.service.StorageService;
import org.apache.cassandra.utils.*;
import org.apache.cassandra.utils.concurrent.Ref;
@@ -349,21 +351,25 @@ class LogTransaction extends
Transactional.AbstractTransactional implements Tran
// must not retain a reference to the SSTableReader, else leak
detection cannot kick in
private final Descriptor desc;
private final long sizeOnDisk;
- private final Tracker tracker;
private final boolean wasNew;
private final Object lock;
private final Ref<LogTransaction> parentRef;
- private final UUID txnId;
+ private final Counter totalDiskSpaceUsed;
public SSTableTidier(SSTableReader referent, boolean wasNew,
LogTransaction parent)
{
this.desc = referent.descriptor;
this.sizeOnDisk = referent.bytesOnDisk();
- this.tracker = parent.tracker;
this.wasNew = wasNew;
this.lock = parent.lock;
this.parentRef = parent.selfRef.tryRef();
- this.txnId = parent.id();
+
+ // While the parent cfs may be dropped in the interim of us taking
a reference to this and using it, at worst
+ // we'll be updating a metric for a now dropped ColumnFamilyStore.
We do not hold a reference to the tracker or
+ // cfs as that would create a strong ref loop and violate our
ability to do leak detection.
+ totalDiskSpaceUsed = parent.tracker != null &&
parent.tracker.cfstore != null ?
+
parent.tracker.cfstore.metric.totalDiskSpaceUsed :
+ null;
if (this.parentRef == null)
throw new IllegalStateException("Transaction already
completed");
@@ -371,7 +377,9 @@ class LogTransaction extends
Transactional.AbstractTransactional implements Tran
public void run()
{
- if (tracker != null && !tracker.isDummy())
+ // While this may be a dummy tracker w/out information in the
metrics table, we attempt to delete regardless
+ // and allow the delete to silently fail if this is an invalid ks
+ cf combination at time of tidy run.
+ if (DatabaseDescriptor.isDaemonInitialized())
SystemKeyspace.clearSSTableReadMeter(desc.ksname, desc.cfname,
desc.generation);
synchronized (lock)
@@ -399,8 +407,11 @@ class LogTransaction extends
Transactional.AbstractTransactional implements Tran
return;
}
- if (tracker != null && tracker.cfstore != null && !wasNew)
- tracker.cfstore.metric.totalDiskSpaceUsed.dec(sizeOnDisk);
+ // It's possible we're the last one's holding a ref to this
metric if it's already been released in the
+ // parent TableMetrics; we run this regardless rather than
holding a ref to that CFS or Tracker and thus
+ // creating a strong ref loop
+ if (DatabaseDescriptor.isDaemonInitialized() &&
totalDiskSpaceUsed != null && !wasNew)
+ totalDiskSpaceUsed.dec(sizeOnDisk);
// release the referent to the parent so that the all
transaction files can be released
parentRef.release();
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
