Janusgraph : Cassandra节俭超时异常

Question

我使用的是内置格式的Janusgraph，后端作为cassandrathrift。这是我给janusgraph和cassandra的财产：

storage.backend=cassandrathrift
storage.cassandra.keyspace=t_graph
storage.cassandra.frame-size-mb=128
storage.hostname=10.XX.XX.XX,20.XX.XX.XX,30.XX.XX.XX

我有一个查询，它找出了一个特定顶点的追随者数。以下是这方面的代码：

public class FollowCountNormal {

    private static JanusGraph graph;
    private static GraphTraversalSource traversalSource;

    public static void main(String[] args) {
        create();
        System.exit(0);
    }

    public static JanusGraph create() {

        graph = JanusGraphFactory.open("/resources/jp.properties");
        traversalSource = graph.traversal();
        getAllEdges();
        return graph;
    }

    static long getAllEdges(){

        try{
            GraphTraversal<Vertex, Vertex> allV = traversalSource.V();
            GraphTraversal<Vertex, Vertex> gt = allV.has("vid", "supernode");
            GraphTraversal<Vertex, Long> c = gt.inE()           
                    .count();
            long l = c.next();
            System.out.println("All edges = "+l);
            graph.tx().commit();
            return l;
        }catch (Exception e) {
            System.out.println("Error while fetching the edges for : ");
            e.printStackTrace();
        }

        return -1;
    }
}

如果顶点有有限数量的传入边，则此代码工作良好(即工作精细是跟随者计数小于100000)。但如果追随者数以百万计，我有以下例外：

造成原因: org.janusgraph.diskstorage.PermanentBackendException:存储后端永久故障在org.janusgraph.diskstorage.cassandra.thrift.CassandraThriftKeyColumnValueStore.convertException(CassandraThriftKeyColumnValueStore.java:263) at org.janusgraph.diskstorage.cassandra.thrift.CassandraThriftKeyColumnValueStore.getNamesSlice(CassandraThriftKeyColumnValueStore.java:162) at org.janusgraph.diskstorage.cassandra.thrift.CassandraThriftKeyColumnValueStore.getNamesSlice(CassandraThriftKeyColumnValueStore.java:105) at org.janusgraph.diskstorage.cassandra.thrift.CassandraThriftKeyColumnValueStore.getSlice(CassandraThriftKeyColumnValueStore.java:94) at org.janusgraph.diskstorage.keycolumnvalue.KCVSProxy.getSlice(KCVSProxy.java:77) at org.janusgraph.diskstorage.keycolumnvalue.cache.ExpirationKCVSCache$2.call(ExpirationKCVSCache.java:100) at org.janusgraph.diskstorage.keycolumnvalue.cache.ExpirationKCVSCache$2.call(ExpirationKCVSCache.java:96) at com.google.common.cache.LocalCache$LocalManualCache$1.load(LocalCache.java:4742) at com.google.common.cache.LocalCache$LoadingValueReference.loadFuture(LocalCache.java:3527) at com.google.common.cache.LocalCache$Segment.loadSync( org.apache.cassandra.thrift.TimedOutException：). 50种常见帧被忽略，原因是: org.apache.cassandra.thrift.Cassandra$multiget_slice_result$multiget_slice_resultStandardScheme.read(Cassandra.java:14696) at org.apache.cassandra.thrift.Cassandra$multiget_slice_result$multiget_slice_resultStandardScheme.read(Cassandra.java:14633) at org.apache.cassandra.thrift.Cassandra$multiget_slice_result.read(Cassandra.java:14559) atorg.apache.thrift.TServiceClient.receiveBase(TServiceClient.java:78) at org.apache.cassandra.thrift.Cassandra$Client.recv_multiget_slice(Cassandra.java:741) at org.apache.cassandra.thrift.Cassandra$Client.multiget_slice(Cassandra.java:725) at org.janusgraph.diskstorage.cassandra.thrift.CassandraThriftKeyColumnValueStore.getNamesSlice(CassandraThriftKeyColumnValueStore.java:143) .省略了58个常见帧

我在网上找了很多这样的帖子：

link1

但对我来说没有解决办法。对如何解决这个问题有什么建议吗？

Answer 1

将存储后端更改为cql和其他与cql相关的属性为我解决了这个问题。以下是我使用的属性：

storage.backend=cql
storage.cql.keyspace=t_graph
storage.cql.read-consistency-level=ONE