[
https://issues.apache.org/jira/browse/LUCENE-10572?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17538198#comment-17538198
]
Uwe Schindler commented on LUCENE-10572:
----------------------------------------
If we have 2 hash tables, we could have one for short terms up to 255 bytes (
for sure could also make the limit smaller, but 255 is the limit to get the 1
byte length encoding), and all longer ones in a separate hash (where also the
comparisons are more expensive).
I am not sure if the additioal complexity is worth to do this.
About changing the hash algorithm: we may add a counter into the hash table to
actually measure how many collisions we have during indexing wikipedia. But
actually when inserting a term already in the hash-table, we get a hash
collision and have to confirm with Array.equals() that the term is already
there. I tend to think that the smaller terms are more often duplicates than
larger ones, so having them in a separate table may be a good idea.
Maybe we should have some statistics during wikipedia indexing:
- how many hash collisions do we have (where term is actually not already in
table)? => this ratio should be low. We can compare hash algorithms for that.
- how many hash collisions do we get because the term is already in table? =>
this is most expensive memory-wise, because hash AND equals have to be
calculated.
- how many inserts of new terms without a collision do we get?
> Can we optimize BytesRefHash?
> -----------------------------
>
> Key: LUCENE-10572
> URL: https://issues.apache.org/jira/browse/LUCENE-10572
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Michael McCandless
> Priority: Major
> Attachments: Screen Shot 2022-05-16 at 10.28.22 AM.png
>
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> I was poking around in our nightly benchmarks
> ([https://home.apache.org/~mikemccand/lucenebench]) and noticed in the JFR
> profiling that the hottest method is this:
> {noformat}
> PERCENT CPU SAMPLES STACK
> 9.28% 53848 org.apache.lucene.util.BytesRefHash#equals()
> at
> org.apache.lucene.util.BytesRefHash#findHash()
> at org.apache.lucene.util.BytesRefHash#add()
> at
> org.apache.lucene.index.TermsHashPerField#add()
> at
> org.apache.lucene.index.IndexingChain$PerField#invert()
> at
> org.apache.lucene.index.IndexingChain#processField()
> at
> org.apache.lucene.index.IndexingChain#processDocument()
> at
> org.apache.lucene.index.DocumentsWriterPerThread#updateDocuments() {noformat}
> This is kinda crazy – comparing if the term to be inserted into the inverted
> index hash equals the term already added to {{BytesRefHash}} is the hottest
> method during nightly benchmarks.
> Discussing offline with [~rcmuir] and [~jpountz] they noticed a few
> questionable things about our current implementation:
> * Why are we using a 1 or 2 byte {{vInt}} to encode the length of the
> inserted term into the hash? Let's just use two bytes always, since IW
> limits term length to 32 K (< 64K that an unsigned short can cover)
> * Why are we doing byte swapping in this deep hotspot using {{VarHandles}}
> (BitUtil.VH_BE_SHORT.get)
> * Is it possible our growth strategy for {{BytesRefHash}} (on rehash) is not
> aggressive enough? Or the initial sizing of the hash is too small?
> * Maybe {{MurmurHash}} is not great (causing too many conflicts, and too
> many {{equals}} calls as a result?) – {{Fnv}} and {{xxhash}} are possible
> "upgrades"?
> * If we stick with {{{}MurmurHash{}}}, why are we using the 32 bit version
> ({{{}murmurhash3_x86_32{}}})?
> * Are we using the JVM's intrinsics to compare multiple bytes in a single
> SIMD instruction ([~rcmuir] is quite sure we are indeed)?
> * [~jpountz] suggested maybe the hash insert is simply memory bound
> * {{TermsHashPerField.writeByte}} is also depressingly slow (~5% of total
> CPU cost)
> I pulled these observations from a recent (5/6/22) profiler output:
> [https://home.apache.org/~mikemccand/lucenebench/2022.05.06.06.33.00.html]
> Maybe we can improve our performance on this crazy hotspot?
> Or maybe this is a "healthy" hotspot and we should leave it be!
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
