@Override public org.apache.hadoop.mapred.InputSplit[] getSplits(JobConf job, int numSplits) throws IOException { Settings settings = HadoopSettingsManager.loadFrom(job); Collection<PartitionDefinition> partitions = RestService.findPartitions(settings, log); EsInputSplit[] splits = new EsInputSplit[partitions.size()]; int index = 0; for (PartitionDefinition part : partitions) { splits[index++] = new EsInputSplit(part); } log.info(String.format("Created [%d] splits", splits.length)); return splits; }
void init(EsInputSplit esSplit, Configuration cfg, Progressable progressable) { // get a copy to override the host/port Settings settings = HadoopSettingsManager.loadFrom(cfg).copy().load(esSplit.getPartition().getSerializedSettings()); if (log.isTraceEnabled()) { log.trace(String.format("Init shard reader from cfg %s", HadoopCfgUtils.asProperties(cfg))); log.trace(String.format("Init shard reader w/ settings %s", settings)); } this.esSplit = esSplit; // initialize mapping/ scroll reader InitializationUtils.setValueReaderIfNotSet(settings, WritableValueReader.class, log); PartitionDefinition part = esSplit.getPartition(); PartitionReader partitionReader = RestService.createReader(settings, part, log); this.scrollReader = partitionReader.scrollReader; this.client = partitionReader.client; this.queryBuilder = partitionReader.queryBuilder; this.progressable = progressable; // in Hadoop-like envs (Spark) the progressable might be null and thus the heart-beat is not needed if (progressable != null) { beat = new HeartBeat(progressable, cfg, settings.getHeartBeatLead(), log); } if (log.isDebugEnabled()) { log.debug(String.format("Initializing RecordReader for [%s]", esSplit)); } }
void init(EsInputSplit esSplit, Configuration cfg, Progressable progressable) { // get a copy to override the host/port Settings settings = HadoopSettingsManager.loadFrom(cfg).copy().load(esSplit.getPartition().getSerializedSettings()); if (log.isTraceEnabled()) { log.trace(String.format("Init shard reader from cfg %s", HadoopCfgUtils.asProperties(cfg))); log.trace(String.format("Init shard reader w/ settings %s", settings)); } this.esSplit = esSplit; // initialize mapping/ scroll reader InitializationUtils.setValueReaderIfNotSet(settings, WritableValueReader.class, log); PartitionDefinition part = esSplit.getPartition(); PartitionReader partitionReader = RestService.createReader(settings, part, log); this.scrollReader = partitionReader.scrollReader; this.client = partitionReader.client; this.queryBuilder = partitionReader.queryBuilder; this.progressable = progressable; // in Hadoop-like envs (Spark) the progressable might be null and thus the heart-beat is not needed if (progressable != null) { beat = new HeartBeat(progressable, cfg, settings.getHeartBeatLead(), log); } if (log.isDebugEnabled()) { log.debug(String.format("Initializing RecordReader for [%s]", esSplit)); } }
void init(EsInputSplit esSplit, Configuration cfg, Progressable progressable) { // get a copy to override the host/port Settings settings = HadoopSettingsManager.loadFrom(cfg).copy().load(esSplit.getPartition().getSerializedSettings()); if (log.isTraceEnabled()) { log.trace(String.format("Init shard reader from cfg %s", HadoopCfgUtils.asProperties(cfg))); log.trace(String.format("Init shard reader w/ settings %s", settings)); } this.esSplit = esSplit; // initialize mapping/ scroll reader InitializationUtils.setValueReaderIfNotSet(settings, WritableValueReader.class, log); PartitionDefinition part = esSplit.getPartition(); PartitionReader partitionReader = RestService.createReader(settings, part, log); this.scrollReader = partitionReader.scrollReader; this.client = partitionReader.client; this.queryBuilder = partitionReader.queryBuilder; this.progressable = progressable; // in Hadoop-like envs (Spark) the progressable might be null and thus the heart-beat is not needed if (progressable != null) { beat = new HeartBeat(progressable, cfg, settings.getHeartBeatLead(), log); } if (log.isDebugEnabled()) { log.debug(String.format("Initializing RecordReader for [%s]", esSplit)); } }
void init(EsInputSplit esSplit, Configuration cfg, Progressable progressable) { // get a copy to override the host/port Settings settings = HadoopSettingsManager.loadFrom(cfg).copy().load(esSplit.getPartition().getSerializedSettings()); if (log.isTraceEnabled()) { log.trace(String.format("Init shard reader from cfg %s", HadoopCfgUtils.asProperties(cfg))); log.trace(String.format("Init shard reader w/ settings %s", settings)); } this.esSplit = esSplit; // initialize mapping/ scroll reader InitializationUtils.setValueReaderIfNotSet(settings, WritableValueReader.class, log); PartitionDefinition part = esSplit.getPartition(); PartitionReader partitionReader = RestService.createReader(settings, part, log); this.scrollReader = partitionReader.scrollReader; this.client = partitionReader.client; this.queryBuilder = partitionReader.queryBuilder; this.progressable = progressable; // in Hadoop-like envs (Spark) the progressable might be null and thus the heart-beat is not needed if (progressable != null) { beat = new HeartBeat(progressable, cfg, settings.getHeartBeatLead(), log); } if (log.isDebugEnabled()) { log.debug(String.format("Initializing RecordReader for [%s]", esSplit)); } }
EsHiveSplit() { this(new EsInputSplit(), null); }
@Override public org.apache.hadoop.mapred.InputSplit[] getSplits(JobConf job, int numSplits) throws IOException { Settings settings = HadoopSettingsManager.loadFrom(job); Collection<PartitionDefinition> partitions = RestService.findPartitions(settings, log); EsInputSplit[] splits = new EsInputSplit[partitions.size()]; int index = 0; for (PartitionDefinition part : partitions) { splits[index++] = new EsInputSplit(part); } log.info(String.format("Created [%d] splits", splits.length)); return splits; }
@Override public org.apache.hadoop.mapred.InputSplit[] getSplits(JobConf job, int numSplits) throws IOException { Settings settings = HadoopSettingsManager.loadFrom(job); Collection<PartitionDefinition> partitions = RestService.findPartitions(settings, log); EsInputSplit[] splits = new EsInputSplit[partitions.size()]; int index = 0; for (PartitionDefinition part : partitions) { splits[index++] = new EsInputSplit(part); } log.info(String.format("Created [%d] splits", splits.length)); return splits; }
@Override public org.apache.hadoop.mapred.InputSplit[] getSplits(JobConf job, int numSplits) throws IOException { Settings settings = HadoopSettingsManager.loadFrom(job); Collection<PartitionDefinition> partitions = RestService.findPartitions(settings, log); EsInputSplit[] splits = new EsInputSplit[partitions.size()]; int index = 0; for (PartitionDefinition part : partitions) { splits[index++] = new EsInputSplit(part); } log.info(String.format("Created [%d] splits", splits.length)); return splits; }
EsHiveSplit() { this(new EsInputSplit(), null); }