[
https://issues.apache.org/jira/browse/ARTEMIS-2513?focusedWorklogId=326067&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-326067
]
ASF GitHub Bot logged work on ARTEMIS-2513:
-------------------------------------------
Author: ASF GitHub Bot
Created on: 10/Oct/19 02:28
Start Date: 10/Oct/19 02:28
Worklog Time Spent: 10m
Work Description: wy96f commented on pull request #2859: ARTEMIS-2513
Large message's copy may be interfered by other threads
URL: https://github.com/apache/activemq-artemis/pull/2859#discussion_r333306857
##########
File path:
artemis-server/src/main/java/org/apache/activemq/artemis/core/persistence/impl/journal/LargeServerMessageImpl.java
##########
@@ -368,52 +368,48 @@ public Message copy(final long newID) {
try {
LargeServerMessage newMessage =
storageManager.createLargeMessage(newID, this);
- boolean originallyOpen = file != null && file.isOpen();
+ //clone a SequentialFile to avoid concurrent access
+ ensureFileExists(false);
+ SequentialFile cloneFile = file.cloneFile();
- validateFile();
-
- byte[] bufferBytes = new byte[100 * 1024];
-
- ByteBuffer buffer = ByteBuffer.wrap(bufferBytes);
+ try {
+ byte[] bufferBytes = new byte[100 * 1024];
- long oldPosition = file.position();
+ ByteBuffer buffer = ByteBuffer.wrap(bufferBytes);
- if (!file.isOpen()) {
- file.open();
- }
- file.position(0);
-
- for (;;) {
- // The buffer is reused...
- // We need to make sure we clear the limits and the buffer before
reusing it
- buffer.clear();
- int bytesRead = file.read(buffer);
-
- byte[] bufferToWrite;
- if (bytesRead <= 0) {
- break;
- } else if (bytesRead == bufferBytes.length &&
!this.storageManager.isReplicated()) {
- // ARTEMIS-1220: We cannot reuse the same buffer if it's
replicated
- // otherwise there could be another thread still using the
buffer on a
- // replication.
- bufferToWrite = bufferBytes;
- } else {
- bufferToWrite = new byte[bytesRead];
- System.arraycopy(bufferBytes, 0, bufferToWrite, 0, bytesRead);
+ if (!cloneFile.isOpen()) {
+ cloneFile.open();
}
- newMessage.addBytes(bufferToWrite);
-
- if (bytesRead < bufferBytes.length) {
- break;
+ cloneFile.position(0);
+
+ for (;;) {
+ // The buffer is reused...
+ // We need to make sure we clear the limits and the buffer
before reusing it
+ buffer.clear();
+ int bytesRead = cloneFile.read(buffer);
+
+ byte[] bufferToWrite;
+ if (bytesRead <= 0) {
+ break;
+ } else if (bytesRead == bufferBytes.length &&
!this.storageManager.isReplicated()) {
+ // ARTEMIS-1220: We cannot reuse the same buffer if it's
replicated
+ // otherwise there could be another thread still using the
buffer on a
+ // replication.
+ bufferToWrite = bufferBytes;
+ } else {
+ bufferToWrite = new byte[bytesRead];
+ System.arraycopy(bufferBytes, 0, bufferToWrite, 0,
bytesRead);
+ }
+
+ newMessage.addBytes(bufferToWrite);
+
+ if (bytesRead < bufferBytes.length) {
+ break;
+ }
}
- }
-
- file.position(oldPosition);
-
- if (!originallyOpen) {
- file.close(false);
- newMessage.getFile().close();
+ } finally {
+ cloneFile.close();
}
Review comment:
@clebertsuconic @gaohoward I don't understand why we only close file when
old file is not originally opened? In your case where other consumers open file
and deliver, the file of new large message might not be closed. This will
result in file leak and maybe data corrupt if broker crashes, wdyt?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 326067)
Time Spent: 1h 20m (was: 1h 10m)
> Large message's copy may be interfered by other threads
> -------------------------------------------------------
>
> Key: ARTEMIS-2513
> URL: https://issues.apache.org/jira/browse/ARTEMIS-2513
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
> Affects Versions: 2.10.1
> Reporter: Howard Gao
> Assignee: Howard Gao
> Priority: Major
> Fix For: 2.11.0
>
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> In LargeMessageImpl.copy(long) it need to open the underlying file in order
> to read and copy bytes into the new copied message. However there is a chance
> that another thread can come in and close the file in the middle, making the
> copy failed with "channel is null" error.
> This is happening in cases where a large message is sent to a jms topic
> (multicast address). During delivery it to multiple subscribers, some
> consumer is doing delivery and closed the underlying file after. Some other
> consumer is rolling back the messages and eventually move it to DLQ (which
> will call the above copy method). So there is a chance this bug being hit on.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
