Configuring Lucene Version Requirements

<luceneMatchVersion>4.4</luceneMatchVersion>

Designing the Schema

Configuring the Heap Size

Set the Java heap size for the Solr Server to at least 16 GB for production environments. For more information on memory requirements, see Guidelines for Deploying Cloudera Search.

General Tips

• Enabling multi-threaded faceting can provide better performance for field faceting. When multi-threaded faceting is enabled, field faceting tasks are completed in a parallel with a thread working on every field faceting task simultaneously. Performance improvements do not occur in all cases, but improvements are likely when all of the following are true:
  • The system uses highly concurrent hardware.
  • Faceting operations apply to large data sets over multiple fields.
  • There is not an unusually high number of queries occurring simultaneously on the system. Systems that are lightly loaded or that are mainly engaged with ingestion and indexing may be helped by multi-threaded faceting; for example, a system ingesting articles and being queried by a researcher. Systems heavily loaded by user queries are less likely to be helped by multi-threaded faceting; for example, an e-commerce site with heavy user-traffic.
    Note: Multi-threaded faceting only applies to field faceting and not to query faceting.

  • Field faceting identifies the number of unique entries for a field. For example, multi-threaded faceting could be used to simultaneously facet for the number of unique entries for the fields, "color" and "size". In such a case, there would be two threads, and each thread would work on faceting one of the two fields.
  • Query faceting identifies the number of unique entries that match a query for a field. For example, query faceting could be used to find the number of unique entries in the "size" field are between 1 and 5. Multi-threaded faceting does not apply to these operations.
  To enable multi-threaded faceting, add facet-threads to queries. For example, to use up to 1000 threads, you might use a query as follows:

  http://localhost:8983/solr/collection1/select?q=*:*&facet=true&fl=id&facet.field=f0_ws&facet.threads=1000

  If facet-threads is omitted or set to 0, faceting is single-threaded. If facet-threads is set to a negative value, such as -1, multi-threaded faceting will use as many threads as there are fields to facet up to the maximum number of threads possible on the system.
• If your environment does not require Near Real Time (NRT), turn off soft auto-commit in solrconfig.xml.
• In most cases, do not change the default batchSize setting of 1000. If you are working with especially large documents, you may consider decreasing the batch size.
• To help identify any garbage collector (GC) issues, enable GC logging in production. The overhead is low and the JVM supports GC log rolling as of 1.6.0_34.
  • The minimum recommended GC logging flags are: -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -XX:+PrintGCDetails.
  • To rotate the GC logs: -Xloggc: -XX:+UseGCLogFileRotation -XX:NumberOfGCLogFiles= -XX:GCLogFileSize=.

    For Cloudera Manager environments, you can set these flags at Solr service > Configuration > Category > Java Configuration Options for Solr Server.

    For unmanaged environments, you can configure Java options by adding or modifying the CATALINA_OPTS environment variable in /etc/default/solr:

    CATALINA_OPTS="-Xmx10g -XX:MaxDirectMemorySize=20g \
    -XX:+UseLargePages -Dsolr.hdfs.blockcache.slab.count=100" \
    -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -XX:+PrintGCDetails

Solr and HDFS - the Block Cache

Cloudera Search enables Solr to store indexes in an HDFS filesystem. To maintain performance, an HDFS block cache has been implemented using Least Recently Used (LRU) semantics. This enables Solr to cache HDFS index files on read and write, storing the portions of the file in JVM direct memory (off heap) by default, or optionally in the JVM heap.

Batch jobs typically do not use the cache, while Solr servers (when serving queries or indexing documents) should. When running indexing using MapReduce, the MR jobs themselves do not use the block cache. Block caching is turned off by default and should be left disabled.

Tuning of this cache is complex and best practices are continually being refined. In general, allocate a cache that is about 10-20% of the amount of memory available on the system. For example, when running HDFS and Solr on a host with 96 GB of memory, allocate 10-20 GB of memory using solr.hdfs.blockcache.slab.count. As index sizes grow you may need to tune this parameter to maintain optimal performance.

  Note: Block cache metrics are currently unavailable.

Configuration

The following parameters control caching. They can be configured at the Solr process level by setting the respective Java system property or by editing solrconfig.xml directly. For more information on setting Java system properties, see Setting Java System Properties for Solr.

If the parameters are set at the collection level (using solrconfig.xml), the first collection loaded by the Solr server takes precedence, and block cache settings in all other collections are ignored. Because you cannot control the order in which collections are loaded, you must make sure to set identical block cache settings in every collection solrconfig.xml. Block cache parameters set at the collection level in solrconfig.xml also take precedence over parameters at the process level.

  Note:

Increasing the direct memory cache size may make it necessary to increase the maximum direct memory size allowed by the JVM. Each Solr slab allocates memory, which is 128 MB by default, as well as allocating some additional direct memory overhead. Therefore, ensure that the MaxDirectMemorySize is set comfortably above the value expected for slabs alone. The amount of additional memory required varies according to multiple factors, but for most cases, setting MaxDirectMemorySize to at least 20-30% more than the total memory configured for slabs is sufficient. Setting MaxDirectMemorySize to the number of slabs multiplied by the slab size does not provide enough memory.

To set MaxDirectMemorySize using Cloudera Manager:

1. Go to the Solr service.
2. Click the Configuration tab.
3. In the Search box, type Java Direct Memory Size of Solr Server in Bytes.
4. Set the new direct memory value.
5. Restart Solr servers after editing the parameter.

To set MaxDirectMemorySize in unmanaged environments:

1. Add -XX:MaxDirectMemorySize=20g to the CATALINA_OPTS environment variable in /etc/default/solr.
2. Restart Solr servers:

   sudo service solr-server restart

Solr HDFS optimizes caching when performing NRT indexing using Lucene's NRTCachingDirectory.

Lucene caches a newly created segment if both of the following conditions are true:

• The segment is the result of a flush or a merge and the estimated size of the merged segment is <= solr.hdfs.nrtcachingdirectory.maxmergesizemb.
• The total cached bytes is <= solr.hdfs.nrtcachingdirectory.maxcachedmb.

The following parameters control NRT caching behavior:

Here is an example of solrconfig.xml with defaults:

 <directoryFactory name="DirectoryFactory">
    <bool name="solr.hdfs.blockcache.enabled">${solr.hdfs.blockcache.enabled:true}</bool>
    <int name="solr.hdfs.blockcache.slab.count">${solr.hdfs.blockcache.slab.count:1}</int>
    <bool name="solr.hdfs.blockcache.direct.memory.allocation">${solr.hdfs.blockcache.direct.memory.allocation:true}</bool>
    <int name="solr.hdfs.blockcache.blocksperbank">${solr.hdfs.blockcache.blocksperbank:16384}</int>
    <bool name="solr.hdfs.blockcache.read.enabled">${solr.hdfs.blockcache.read.enabled:true}</bool>
    <bool name="solr.hdfs.blockcache.write.enabled">${solr.hdfs.blockcache.write.enabled:true}</bool>
    <bool name="solr.hdfs.nrtcachingdirectory.enable">${solr.hdfs.nrtcachingdirectory.enable:true}</bool>
    <int name="solr.hdfs.nrtcachingdirectory.maxmergesizemb">${solr.hdfs.nrtcachingdirectory.maxmergesizemb:16}</int>
    <int name="solr.hdfs.nrtcachingdirectory.maxcachedmb">${solr.hdfs.nrtcachingdirectory.maxcachedmb:192}</int>
</directoryFactory>

The following example illustrates passing Java options by editing the /etc/default/solr or /opt/cloudera/parcels/CDH-*/etc/default/solr configuration file:

CATALINA_OPTS="-Xmx10g -XX:MaxDirectMemorySize=20g -XX:+UseLargePages \
-Dsolr.hdfs.blockcache.slab.count=100"

For better performance, Cloudera recommends setting the Linux swap space on all Solr server hosts as shown below:

• Minimize swappiness:

  sudo sysctl vm.swappiness=1

• Disable swap space until next reboot:

  sudo swapoff -a

  Note: Cloudera previously recommended setting vm.swappiness to 0. However, a change in Linux kernel 3.5-rc1 (fe35004f), can lead to frequent out of memory (OOM) errors. For details, see Evan Klitzke's blog post. This commit was backported to RHEL / CentOS 6.4 and Ubuntu 12.04 LTS (Long Term Support).

Garbage Collection

Choose different garbage collection options for best performance in different environments. Some garbage collection options typically chosen include:

• Concurrent low pause collector: Use this collector in most cases. This collector attempts to minimize "Stop the World" events. Avoiding these events can reduce connection timeouts, such as with ZooKeeper, and may improve user experience. This collector is enabled using the Java system property -XX:+UseConcMarkSweepGC.
• Throughput collector: Consider this collector if raw throughput is more important than user experience. This collector typically uses more "Stop the World" events so this may negatively affect user experience and connection timeouts such as ZooKeeper heartbeats. This collector is enabled using the Java system property -XX:+UseParallelGC. If UseParallelGC "Stop the World" events create problems, such as ZooKeeper timeouts, consider using the UseParNewGC collector as an alternative collector with similar throughput benefits.

For information on setting Java system properties, see Setting Java System Properties for Solr.

You can also affect garbage collection behavior by increasing the Eden space to accommodate new objects. With additional Eden space, garbage collection does not need to run as frequently on new objects.

Replication

You can adjust the degree to which different data is replicated.

Replication Settings

  Note: Do not adjust HDFS replication settings for Solr in most cases.

To adjust the Solr replication factor for index files stored in HDFS:

• Cloudera Manager:
  1. Go to Solr service > Configuration > Category > Advanced.
  2. Click the plus sign next to Solr Service Advanced Configuration Snippet (Safety Valve) for hdfs-site.xml to add a new property with the following values:

     Name: dfs.replication

     Value: 2

  3. Click Save Changes.
  4. Restart the Solr service (Solr service > Actions > Restart).
• Unmanaged:
  1. Configure the solr.hdfs.confdir system property to refer to the Solr HDFS configuration files. Typically the value is /etc/solrhdfs/. For information on setting Java system properties, see Setting Java System Properties for Solr.
  2. Set the DFS replication value in the HDFS configuration file at the location you specified in the previous step. For example, to set the replication value to 2, you would change the dfs.replication setting as follows:

     <property>
       <name>dfs.replication<name>
       <value>2<value>
     <property>

  3. Restart the Solr service:

     sudo service solr-server restart

Replicas

If you have sufficient additional hardware, add more replicas for a linear boost of query throughput. Note that adding replicas may slow write performance on the first replica, but otherwise this should have minimal negative consequences.

Transaction Log Replication

Beginning with CDH 5.4.1, Search supports configurable transaction log replication levels for replication logs stored in HDFS. Cloudera recommends leaving the value unchanged at 3 or, barring that, setting it to at least 2.

 <updateHandler class="solr.DirectUpdateHandler2">

    <!-- Enables a transaction log, used for real-time get, durability, and
         and solr cloud replica recovery.  The log can grow as big as
         uncommitted changes to the index, so use of a hard autoCommit
         is recommended (see below).
         "dir" - the target directory for transaction logs, defaults to the
                solr data directory.  -->
    <updateLog>
      <str name="dir">${solr.ulog.dir:}</str>
      <int name="tlogDfsReplication">${solr.ulog.tlogDfsReplication:3}</int>
      <int name="numVersionBuckets">${solr.ulog.numVersionBuckets:65536}</int>
    </updateLog>

The default replication level is 3. For clusters with fewer than three DataNodes (such as proof-of-concept clusters), reduce this number to the amount of DataNodes in the cluster. Changing the replication level only applies to new transaction logs.

Initial testing shows no significant performance regression for common use cases.

Shards

In some cases, oversharding can help improve performance including intake speed. If your environment includes massively parallel hardware and you want to use these available resources, consider oversharding. You might increase the number of replicas per host from 1 to 2 or 3. Making such changes creates complex interactions, so you should continue to monitor your system's performance to ensure that the benefits of oversharding do not outweigh the costs.

Commits

Changing commit values may improve performance in some situation. These changes result in tradeoffs and may not be beneficial in all cases.

• For hard commit values, the default value of 60000 (60 seconds) is typically effective, though changing this value to 120 seconds may improve performance in some cases. Note that setting this value to higher values, such as 600 seconds may result in undesirable performance tradeoffs.
• Consider increasing the auto-commit value from 15000 (15 seconds) to 120000 (120 seconds).
• Enable soft commits and set the value to the largest value that meets your requirements. The default value of 1000 (1 second) is too aggressive for some environments.