SOLR-6481: CLUSTERSTATUS action should consult /live_nodes when reporting the state of a replica

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1626818 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Timothy Potter 2014-09-22 16:17:16 +00:00
parent e8e49968c9
commit 05a55f9786
2 changed files with 39 additions and 1 deletions

View File

@ -186,6 +186,9 @@ Bug Fixes
* SOLR-6507: Fixed several bugs involving stats.field used with local params (hossman)
* SOLR-6481: CLUSTERSTATUS should check if the node hosting a replica is live when
reporting replica status (Timothy Potter)
Other Changes
----------------------

View File

@ -40,6 +40,7 @@ import java.util.Collection;
import java.util.Collections;
import java.util.HashMap;
import java.util.HashSet;
import java.util.Iterator;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Locale;
@ -795,6 +796,10 @@ public class OverseerCollectionProcessor implements Runnable, Closeable {
}
}
List<String> liveNodes = zkStateReader.getZkClient().getChildren(ZkStateReader.LIVE_NODES_ZKNODE, null, true);
// now we need to walk the collectionProps tree to cross-check replica state with live nodes
crossCheckReplicaStateWithLiveNodes(liveNodes, collectionProps);
NamedList<Object> clusterStatus = new SimpleOrderedMap<>();
clusterStatus.add("collections", collectionProps);
@ -816,12 +821,42 @@ public class OverseerCollectionProcessor implements Runnable, Closeable {
}
// add live_nodes
List<String> liveNodes = zkStateReader.getZkClient().getChildren(ZkStateReader.LIVE_NODES_ZKNODE, null, true);
clusterStatus.add("live_nodes", liveNodes);
results.add("cluster", clusterStatus);
}
/**
* Walks the tree of collection status to verify that any replicas not reporting a "down" status is
* on a live node, if any replicas reporting their status as "active" but the node is not live is
* marked as "down"; used by CLUSTERSTATUS.
* @param liveNodes List of currently live node names.
* @param collectionProps Map of collection status information pulled directly from ZooKeeper.
*/
protected void crossCheckReplicaStateWithLiveNodes(List<String> liveNodes, NamedList<Object> collectionProps) {
Iterator<Map.Entry<String,Object>> colls = collectionProps.iterator();
while (colls.hasNext()) {
Map.Entry<String,Object> next = colls.next();
Map<String,Object> collMap = (Map<String,Object>)next.getValue();
Map<String,Object> shards = (Map<String,Object>)collMap.get("shards");
for (Object nextShard : shards.values()) {
Map<String,Object> shardMap = (Map<String,Object>)nextShard;
Map<String,Object> replicas = (Map<String,Object>)shardMap.get("replicas");
for (Object nextReplica : replicas.values()) {
Map<String,Object> replicaMap = (Map<String,Object>)nextReplica;
if (!ZkStateReader.DOWN.equals(replicaMap.get(ZkStateReader.STATE_PROP))) {
// not down, so verify the node is live
String node_name = (String)replicaMap.get(ZkStateReader.NODE_NAME_PROP);
if (!liveNodes.contains(node_name)) {
// node is not live, so this replica is actually down
replicaMap.put(ZkStateReader.STATE_PROP, ZkStateReader.DOWN);
}
}
}
}
}
}
/**
* Get collection status from cluster state.
* Can return collection status by given shard name.