[
https://issues.apache.org/jira/browse/SOLR-12878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16673334#comment-16673334
]
Tim Underwood commented on SOLR-12878:
--------------------------------------
There are 2 other LeafReader implementations not caching their FieldInfos:
* org.apache.solr.search.CollapsingQParserPlugin.ReaderWrapper
* org.apache.solr.handler.component.ExpandComponent.ReaderWrapper
Both have code that is basically this:
{code:java}
@SuppressWarnings("resource") LeafReader uninvertingReader =
UninvertingReader.wrap(
new ReaderWrapper(searcher.getSlowAtomicReader(), field),
Collections.singletonMap(field, UninvertingReader.Type.SORTED)::get);
{code}
Note the searcher.getSlowAtomicReader() call which means moving the caching of
FieldInfos into SlowCompositeReaderWrapper will help in this case!
Here is part of the code from UninvertingReader.wrap:
{code:java}
public static LeafReader wrap(LeafReader in, Function<String, Type> mapping) {
boolean wrap = false;
// Calculate a new FieldInfos that has DocValuesType where we didn't before
ArrayList<FieldInfo> newFieldInfos = new
ArrayList<>(in.getFieldInfos().size());
for (FieldInfo fi : in.getFieldInfos()) {
DocValuesType type = fi.getDocValuesType();
{code}
Note the 2 immediate calls to in.getFieldInfos(). So... I'll add the FieldInfo
caching to both of those ReaderWrapper classes since it would help.
> FacetFieldProcessorByHashDV is reconstructing FieldInfos on every
> instantiation
> -------------------------------------------------------------------------------
>
> Key: SOLR-12878
> URL: https://issues.apache.org/jira/browse/SOLR-12878
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Components: Facet Module
> Affects Versions: 7.5
> Reporter: Tim Underwood
> Priority: Major
> Labels: performance
> Fix For: 7.6, master (8.0)
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> The FacetFieldProcessorByHashDV constructor is currently calling:
> {noformat}
> FieldInfo fieldInfo =
> fcontext.searcher.getSlowAtomicReader().getFieldInfos().fieldInfo(sf.getName());
> {noformat}
> Which is reconstructing FieldInfos each time. Simply switching it to:
> {noformat}
> FieldInfo fieldInfo =
> fcontext.searcher.getFieldInfos().fieldInfo(sf.getName());
> {noformat}
>
> causes it to use the cached version of FieldInfos in the SolrIndexSearcher.
> On my index the FacetFieldProcessorByHashDV is 2-3 times slower than the
> legacy facets without this fix.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
