[
https://issues.apache.org/jira/browse/CASSANDRA-6283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13905520#comment-13905520
]
Joshua McKenzie commented on CASSANDRA-6283:
--------------------------------------------
There's pros and cons to either side of the debate - there's some benefits to
being able to change file handles while in use such as being able to run
updates, file recovery, etc. Apparently this is a throwback in OS design to
MS-DOS 3.3 as far as file-level locking is concerned.
Coming from a C++ background myself I know what you mean w/smart pointers and
RAII-type resource management, but I believe the problem we're running into
here is language independent. We have atomic reference-counting
implementations in the SSTableReaders to allow multiple concurrent read-only
access to the data structures which would be a complicated implementation
regardless of language choice.
I mentioned process explorer not seeing the file handle lock because I found it
an oddity - my expectation is that anything from Sysinternals and Russinovich
is bullet-proof, so I'm wondering how the OS got into a state where a file is
locked yet I can't query the process that has said lock, even though stopping
the JVM clearly released it.
> Windows 7 data files keept open / can't be deleted after compaction.
> --------------------------------------------------------------------
>
> Key: CASSANDRA-6283
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6283
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Environment: Windows 7 (32) / Java 1.7.0.45
> Reporter: Andreas Schnitzerling
> Assignee: Joshua McKenzie
> Labels: compaction
> Fix For: 2.0.6
>
> Attachments: leakdetect.patch, screenshot-1.jpg, system.log
>
>
> Files cannot be deleted, patch CASSANDRA-5383 (Win7 deleting problem) doesn't
> help on Win-7 on Cassandra 2.0.2. Even 2.1 Snapshot is not running. The cause
> is: Opened file handles seem to be lost and not closed properly. Win 7
> blames, that another process is still using the file (but its obviously
> cassandra). Only restart of the server makes the files deleted. But after
> heavy using (changes) of tables, there are about 24K files in the data folder
> (instead of 35 after every restart) and Cassandra crashes. I experiminted and
> I found out, that a finalizer fixes the problem. So after GC the files will
> be deleted (not optimal, but working fine). It runs now 2 days continously
> without problem. Possible fix/test:
> I wrote the following finalizer at the end of class
> org.apache.cassandra.io.util.RandomAccessReader:
> {code:title=RandomAccessReader.java|borderStyle=solid}
> @Override
> protected void finalize() throws Throwable {
> deallocate();
> super.finalize();
> }
> {code}
> Can somebody test / develop / patch it? Thx.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
