Apache9 commented on a change in pull request #2584: URL: https://github.com/apache/hbase/pull/2584#discussion_r514232329
########## File path: hbase-client/src/main/java/org/apache/hadoop/hbase/client/CatalogReplicaLoadBalanceReplicaSimpleSelector.java ########## @@ -0,0 +1,273 @@ +/** + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.hadoop.hbase.client; + +import static org.apache.hadoop.hbase.client.ConnectionUtils.isEmptyStopRow; +import static org.apache.hadoop.hbase.util.Bytes.BYTES_COMPARATOR; +import static org.apache.hadoop.hbase.util.ConcurrentMapUtils.computeIfAbsent; +import java.io.IOException; +import java.util.Iterator; +import java.util.Map; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.ConcurrentMap; +import java.util.concurrent.ConcurrentNavigableMap; +import java.util.concurrent.ConcurrentSkipListMap; +import java.util.concurrent.ThreadLocalRandom; +import org.apache.commons.lang3.builder.ToStringBuilder; +import org.apache.hadoop.hbase.HRegionLocation; +import org.apache.hadoop.hbase.ScheduledChore; +import org.apache.hadoop.hbase.Stoppable; +import org.apache.hadoop.hbase.TableName; +import org.apache.hadoop.hbase.util.Bytes; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +/** + * CatalogReplicaLoadBalanceReplicaSimpleSelector implements a simple catalog replica load balancing + * algorithm. It maintains a stale location cache for each table. Whenever client looks up location, + * it first check if the row is the stale location cache. If yes, the location from + * catalog replica is stale, it will go to the primary region to look up update-to-date location; + * otherwise, it will randomly pick up a replica region for lookup. When clients receive + * RegionNotServedException from region servers, it will add these region locations to the stale + * location cache. The stale cache will be cleaned up periodically by a chore. + * + * It follows a simple algorithm to choose a replica to go: + * + * 1. If there is no stale location entry for rows it looks up, it will randomly + * pick a replica region to do lookup. + * 2. If the location from the replica region is stale, client gets RegionNotServedException + * from region server, in this case, it will create StaleLocationCacheEntry in + * CatalogReplicaLoadBalanceReplicaSimpleSelector. + * 3. When client tries to do location lookup, it checks StaleLocationCache first for rows it + * tries to lookup, if entry exists, it will go with primary meta region to do lookup; + * otherwise, it will follow step 1. + * 4. A chore will periodically run to clean up cache entries in the StaleLocationCache. + */ +class CatalogReplicaLoadBalanceReplicaSimpleSelector implements + CatalogReplicaLoadBalanceReplicaSelector { + private static final Logger LOG = + LoggerFactory.getLogger(CatalogReplicaLoadBalanceReplicaSimpleSelector.class); + private final long STALE_CACHE_TIMEOUT_IN_MILLISECONDS = 3000; // 3 seconds + private final int STALE_CACHE_CLEAN_CHORE_INTERVAL = 1500; // 1.5 seconds + + /** + * StaleLocationCacheEntry is the entry when a stale location is reported by an client. + */ + private static final class StaleLocationCacheEntry { + // replica id where the stale location comes from. + private int fromReplicaId; + + // timestamp in milliseconds + private long timestamp; + + private byte[] endKey; + + StaleLocationCacheEntry(final int metaReplicaId, final byte[] endKey) { + this.fromReplicaId = metaReplicaId; + this.endKey = endKey; + timestamp = System.currentTimeMillis(); + } + + public byte[] getEndKey() { + return this.endKey; + } + + public int getFromReplicaId() { + return this.fromReplicaId; + } + public long getTimestamp() { + return this.timestamp; + } + + @Override + public String toString() { + return new ToStringBuilder(this) + .append("endKey", endKey) + .append("fromReplicaId", fromReplicaId) + .append("timestamp", timestamp) + .toString(); + } + } + + private static final class StaleTableCache { + private final ConcurrentNavigableMap<byte[], StaleLocationCacheEntry> cache = + new ConcurrentSkipListMap<>(BYTES_COMPARATOR); + } + + private final ConcurrentMap<TableName, StaleTableCache> staleCache; + private int numOfReplicas; + private final AsyncConnectionImpl conn; + private TableName tableName; + + CatalogReplicaLoadBalanceReplicaSimpleSelector(TableName tableName, AsyncConnectionImpl conn) { + staleCache = new ConcurrentHashMap<>(); + this.conn = conn; + this.tableName = tableName; + + // This numOfReplicas is going to be lazy initialized. + this.numOfReplicas = -1; + // Start connection's chore service in case. + this.conn.startChoreService(); + this.conn.getChoreService().scheduleChore(getCacheCleanupChore(this)); + } + + /** + * When a client runs into RegionNotServingException, it will call this method to + * update Selector's internal state. + * @param loc the location which causes exception. + * @param fromReplicaId the replica id where the stale location comes from. + */ + public void notifyOnError(HRegionLocation loc, int fromReplicaId) { + StaleTableCache tableCache = + computeIfAbsent(staleCache, loc.getRegion().getTable(), StaleTableCache::new); + byte[] startKey = loc.getRegion().getStartKey(); + tableCache.cache.putIfAbsent(startKey, + new StaleLocationCacheEntry(fromReplicaId, loc.getRegion().getEndKey())); + LOG.debug("Add entry to stale cache for table {} with startKey {}, {}", + loc.getRegion().getTable(), startKey, loc.getRegion().getEndKey()); + } + + private int getRandomNonPrimaryReplicaId() { + if (numOfReplicas < 0) { + try { + Admin admin = conn.toConnection().getAdmin(); + TableDescriptor tableDescriptor = admin.getDescriptor(tableName); + this.numOfReplicas = tableDescriptor.getRegionReplication(); + } catch (IOException ioe) { + LOG.error("Failed to get table {}'s region replication, exception ", tableName, ioe); + this.numOfReplicas = 1; + } + } + // In case of no replica configured, return the primary region id. + if (this.numOfReplicas <= 1) { + return RegionInfo.DEFAULT_REPLICA_ID; + } + return 1 + ThreadLocalRandom.current().nextInt(this.numOfReplicas - 1); + } + + /** + * When it looks up a location, it will call this method to find a replica replica region to go. + * For a normal case, > 99% of region locations from catalog/meta replica will be up to date. + * In extreme cases such as region server crashes, it will depends on how fast replication + * catches up. + * + * @param tablename table name it looks up + * @param row key it looks up. + * @param locateType locateType, Only BEFORE and CURRENT will be passed in. + * @return catalog replica id + */ + public int selectReplica(final TableName tablename, final byte[] row, + final RegionLocateType locateType) { + StaleTableCache tableCache = staleCache.get(tablename); + + // If there is no entry in StaleCache, select a random replica id. + if (tableCache == null) { + return getRandomNonPrimaryReplicaId(); + } + + Map.Entry<byte[], StaleLocationCacheEntry> entry; + boolean isEmptyStopRow = isEmptyStopRow(row); + // Only BEFORE and CURRENT are passed in. + if (locateType == RegionLocateType.BEFORE) { + entry = isEmptyStopRow ? tableCache.cache.lastEntry() : tableCache.cache.lowerEntry(row); + } else { + entry = tableCache.cache.floorEntry(row); + } + + // It is not in the stale cache, return a random replica id. + if (entry == null) { + return getRandomNonPrimaryReplicaId(); + } + + // Check if the entry times out. + if ((System.currentTimeMillis() - entry.getValue().getTimestamp()) >= + STALE_CACHE_TIMEOUT_IN_MILLISECONDS) { + LOG.debug("Entry for table {} with startKey {}, {} times out", tablename, entry.getKey(), + entry); + tableCache.cache.remove(entry.getKey()); + return getRandomNonPrimaryReplicaId(); + } + + byte[] endKey = entry.getValue().getEndKey(); + + // The following logic is borrowed from AsyncNonMetaRegionLocator. + if (isEmptyStopRow(endKey)) { Review comment: Please add more comments to explain the logic here. I think you should at least say this: "Check the timeout first as comparing long is faster than comparing byte arrays(in most cases), and it could remove the stale entries more quickly. It is not a problem if this is not the entry for us, as if it times out, we will go to random selection, if it does not time out, we will go to the next check and know it is not for us, then we will also go to random selection." ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
