public class DataStreamerCacheUpdaters extends Object
| Modifier and Type | Class and Description |
|---|---|
static interface |
DataStreamerCacheUpdaters.InternalUpdater
Marker interface for updaters which do not need to unwrap cache objects.
|
| Constructor and Description |
|---|
DataStreamerCacheUpdaters() |
| Modifier and Type | Method and Description |
|---|---|
static <K,V> StreamReceiver<K,V> |
batched()
Updates cache using batched methods
IgniteCache.putAll(Map)and
IgniteCache.removeAll(). |
static <K extends Comparable<?>,V> |
batchedSorted()
Updates cache using batched methods
IgniteCache.putAll(Map) and
IgniteCache.removeAll(Set). |
static <K,V> StreamReceiver<K,V> |
individual()
Updates cache using independent
IgniteCache.put(Object, Object)and
IgniteCache.remove(Object) operations. |
protected static <K,V> void |
updateAll(IgniteCache<K,V> cache,
Set<K> rmvCol,
Map<K,V> putMap)
Updates cache.
|
public static <K,V> StreamReceiver<K,V> individual()
IgniteCache.put(Object, Object)and
IgniteCache.remove(Object) operations. Thus it is safe from deadlocks but performance
is not the best.public static <K,V> StreamReceiver<K,V> batched()
IgniteCache.putAll(Map)and
IgniteCache.removeAll(). Can cause deadlocks if the same keys are getting
updated concurrently. Performance is generally better than in individual().public static <K extends Comparable<?>,V> StreamReceiver<K,V> batchedSorted()
IgniteCache.putAll(Map) and
IgniteCache.removeAll(Set). Keys are sorted in natural order and if all updates
use the same rule deadlock can not happen. Performance is generally better than in individual().protected static <K,V> void updateAll(IgniteCache<K,V> cache, @Nullable Set<K> rmvCol, Map<K,V> putMap)
cache - Cache.rmvCol - Keys to remove.putMap - Entries to put.IgniteException - If failed.
Follow @ApacheIgnite
Ignite Fabric : ver. 1.1.0-incubating Release Date : May 20 2015