My project requires me to preload data from a database on startup and I am able to do so by using "preload=true" attribute on persistence. The database I am querying contains half million entries and so I am not supposed to query more than once per node. After several nodes have been started, I noticed that some nodes query the database for all data. Is there any way I can configure Infinispan to stop querying for all data once it has preloaded the entries?

Test config I am using to debug the problem, using a dummy persistence implementation,

<?xml version="1.0" encoding="UTF-8"?>

<infinispan>

  <cache-container default-cache="default" name="cacheContainer" statistics="true">

  <transport cluster="cluster" />

  <distributed-cache-configuration

  name="default-config"

  mode="SYNC"

  statistics="true"

  >

  <persistence>

  <store

  fetch-state="false"

  preload="true"

  purge="false"

  read-only="false"

  class="infinispan.persistence.DummyMapPersistence">

  </store>

  </persistence>

  </distributed-cache-configuration>

  <distributed-cache

  name="default"

  configuration="default-config">

  </distributed-cache>

  </cache-container>

</infinispan>

public class DummyMapPersistence<K, V> implements AdvancedLoadWriteStore<K, V> {

  @Override

  public void init(InitializationContext ctx) {

  log("init");

  }

  private static void log(String msg) {

  System.out.println("[DumbMapPersistence] " + msg);

  }

  @Override

  public MarshalledEntry<K, V> load(Object key) {

  log("load()");

  }

  @Override

  public boolean contains(Object key) {

  }

  @Override

  public void start() {

  }

  @Override

  public void stop() {

  }

  @Override

  public void write(MarshalledEntry<? extends K, ? extends V> entry) {

  log("write()");

  }

  @Override

  public boolean delete(Object key) {

  }

  @Override

  public void process(KeyFilter<? super K> filter,

  org.infinispan.persistence.spi.AdvancedCacheLoader.CacheLoaderTask<K, V> task, Executor executor,

  boolean fetchValue, boolean fetchMetadata) {

  log("process()");

}

  @Override

  public int size() {

  log("size()");

  return plainDatastore.size();

  }

  @Override

  public void clear() {

  log("clear()");

  plainDatastore.clear();

  }

  @Override

  public void purge(Executor threadPool,

  org.infinispan.persistence.spi.AdvancedCacheWriter.PurgeListener<? super K> listener) {

  log("purge()");

}

}

The main:

public class PersistencePreloadTest {

  public static void main(String[] args) throws Exception {

  DefaultCacheManager cacheManager = new DefaultCacheManager("infinispan.xml");

  Cache<String, Object> cache = cacheManager.getCache("default");

  System.out.println("-- Datagrid started --");

  System.out.println("-- process() should not appear after this --");

  }

}

After launching serveral nodes with this same configuration, I find that AdvancedLoadWriteStore::process() is called on some of the nodes. Is there any way to get around this ?

Thanks in advance