Mailing-List: contact dev-help@clerezza.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@clerezza.apache.org
Received-SPF: pass (nike.apache.org: local policy includes SPF record at
 spf.trusted-forwarder.org)
Message-ID: <52988920.2060206@xup.nl>
Date: Fri, 29 Nov 2013 13:31:28 +0100
From: Minto van der Sluis <minto@xup.nl>
Organization: Xup BV
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64;
 rv:24.0) Gecko/20100101 Thunderbird/24.1.1
MIME-Version: 1.0
To: dev@clerezza.apache.org
Subject: Re: Is Clerezza leaking memory?
References: <52974275.5050401@xup.nl> <529852BB.4080008@apache.org>
In-Reply-To: <529852BB.4080008@apache.org>
Content-Type: multipart/alternative;
 boundary="------------050401080202070609020502"

--------------050401080202070609020502
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Andy Seaborne schreef op 29-11-2013 9:39:
> On 28/11/13 13:17, Minto van der Sluis wrote:
>> Hi,
>>
>> I just ran into some peculiar behavior.
>>
>> For my current project I have to import 633 files each containing
>> approx 20 MB of xml data (a total of 13 GB). When importing this data
>> into a single graph I hit an out of memory exception on the 7th file.
>>
>> Looking at the heap I noticed that after restarting the application I
>> could load a few more files. So I started looking for the bundle that
>> consumed all the memory. It happened to be the Clerezza TDB Storage
>> provider. See the following image (GC = garbage collection):
>>
>>
>>
>>
>> Looking more closely I noticed that Apache Jena is able to close a
>> graph (graph.close()) But Clerezza is not using this feature and is
>> keeping the graph open all the time.
>
> Jena graphs backed by TDB are simply views of the dataset - they don't
> have any state associated with them directly.  If the reference become
> inaccessible, GC should clean up.
Hi Andy,

The problem, as far as I can tell, is not in Jena TDB itself. The Jena
TDB bundle is still active/running. Only the Clerezza TDB Provider
bundle is stopped (by me). Like my image shows a normal GC does not
release all of the memory. Only after stopping the Clerezza TDB Provider
memory allocated for importing is release. Because of stopping this
particular bundle all jena datastructures become inaccessible and
eligible for GC. Just like the image shows.

My reasoning is that since the Clerezza TDB Provider has a map with weak
references to Jena models these references are never properly garbage
collected. Since I use the same graph all the time all data gets
accumulated and resulting in out of memory. Looking at a memory dump,
most space is occupied by byte arrays containing the imported data.

I use a nasty hack to prevent this dreaded out of memory. After every
import I restart the Clerezza TDB Provider bundle programmatically (hail
OSGI for I wouldn't know how to do this without OSGI). Like this I have
been able to import more that 300 files in a row (still running).

Regards,

Minto


--------------050401080202070609020502
Content-Type: multipart/related;
 boundary="------------060801040104010809070400"


--------------060801040104010809070400
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">Andy Seaborne schreef op 29-11-2013
      9:39:<br>
    </div>
    <blockquote cite="mid:529852BB.4080008@apache.org" type="cite">
      <meta content="text/html; charset=ISO-8859-1"
        http-equiv="Content-Type">
      <div class="moz-cite-prefix">On 28/11/13 13:17, Minto van der
        Sluis wrote:<br>
      </div>
      <blockquote cite="mid:52974275.5050401@xup.nl" type="cite">
        <meta http-equiv="content-type" content="text/html;
          charset=ISO-8859-1">
        Hi,<br>
        <br>
        I just ran into some peculiar behavior.<br>
        <br>
        For my current project I have to import 633 files each
        containing approx 20 MB of xml data (a total of 13 GB). When
        importing this data into a single graph I hit an out of memory
        exception on the 7th file.<br>
        <br>
        Looking at the heap I noticed that after restarting the
        application I could load a few more files. So I started looking
        for the bundle that consumed all the memory. It happened to be
        the Clerezza TDB Storage provider. See the following image (GC =
        garbage collection):<br>
        <br>
        <br>
        <br>
        <img src="cid:part1.07000307.04020804@xup.nl" alt=""><br>
        Looking more closely I noticed that Apache Jena is able to close
        a graph (graph.close()) But Clerezza is not using this feature
        and is keeping the graph open all the time.<br>
      </blockquote>
      <br>
      Jena graphs backed by TDB are simply views of the dataset - they
      don't have any state associated with them directly.&nbsp; If the
      reference become inaccessible, GC should clean up.</blockquote>
    Hi Andy,<br>
    <br>
    The problem, as far as I can tell, is not in Jena TDB itself. The
    Jena TDB bundle is still active/running. Only the Clerezza TDB
    Provider bundle is stopped (by me). Like my image shows a normal GC
    does not release all of the memory. Only after stopping the Clerezza
    TDB Provider memory allocated for importing is release. Because of
    stopping this particular bundle all jena datastructures become
    inaccessible and eligible for GC. Just like the image shows.<br>
    <br>
    My reasoning is that since the Clerezza TDB Provider has a map with
    weak references to Jena models these references are never properly
    garbage collected. Since I use the same graph all the time all data
    gets accumulated and resulting in out of memory. Looking at a memory
    dump, most space is occupied by byte arrays containing the imported
    data.<br>
    <br>
    I use a nasty hack to prevent this dreaded out of memory. After
    every import I restart the Clerezza TDB Provider bundle
    programmatically (hail OSGI for I wouldn't know how to do this
    without OSGI). Like this I have been able to import more that 300
    files in a row (still running). <br>
    <br>
    Regards,<br>
    <br>
    Minto<br>
    <br>
    <br>
    <br>
  </body>
</html>

--------------060801040104010809070400
Content-Type: image/png
Content-Transfer-Encoding: base64
Content-ID: <part1.07000307.04020804@xup.nl>
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--------------060801040104010809070400--

--------------050401080202070609020502--