Why do you create a new GraphDatabaseService for each of your 1.000 queries? 
Can’t you create one and reuse it?

The “knowledgeGraph” is only created once, as far as I can see, so even though 
pagecache_memory is the right setting, you are only applying it to ~0.1% of the 
databases you start.

--
Chris Vest
System Engineer, Neo Technology
[ skype: mr.chrisvest, twitter: chvest ]


> On 25 Mar 2016, at 07:25, Mohammad Hossain Namaki <[email protected]> wrote:
> 
> Hi all,
> I'm using neo4j.2.3.0.
> I'm running some experiments (over ~1000 graph queries) on the Neo4j Java 
> API. I don't use the cypher for that. So, I have my own algorithm. After 
> running around 100/200 queries, I'm getting the "GC Overhead Limit Exceeded". 
> I think there is something wrong with 
> "org.neo4j.io.pagecache.impl.muninn.MuninnPageCache". 
> 
> Is there anyway to bound this cache size in Java API? or disable it?
> 
> I've already used this:
> knowledgeGraph = new 
> GraphDatabaseFactory().newEmbeddedDatabaseBuilder(MODELGRAPH_DB_PATH)
> 
> .setConfig(GraphDatabaseSettings.pagecache_memory, "500M").newGraphDatabase();
> 
> but it didn't work.
> 
> 
> 
> Every time, I open a nested transaction for both query/data graphs and I 
> success/close them after finishing each query.
> 
> 
> 
> I've used -Xmx12g -Xms12g as VM arguments and my machine has 16GB memory. The 
> data graph itself has 4.00 GB volume on the disk.
> 
> 
> 
> Could you please help me on that?
> 
> 
> 
> Thanks
> 
> 
> 
> 
> 
> 
> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Neo4j" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected] 
> <mailto:[email protected]>.
> For more options, visit https://groups.google.com/d/optout 
> <https://groups.google.com/d/optout>.

-- 
You received this message because you are subscribed to the Google Groups 
"Neo4j" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to