Andrew Hankinson created SOLR-17454:
---------------------------------------
Summary: ERROR message in logs with multithreaded searches
Key: SOLR-17454
URL: https://issues.apache.org/jira/browse/SOLR-17454
Project: Solr
Issue Type: Bug
Security Level: Public (Default Security Level. Issues are Public)
Components: search
Affects Versions: 9.7
Environment: Solr 9.7.0
Ubuntu 22.04 LTS
Reporter: Andrew Hankinson
I sent a message to the Users mailing list about this, but received no
response. However, I think it is still a problem.
When searching with 9.7.0 and enabled multithreaded search, I now get an ERROR
message in my logs:
{code:java}
2024-09-16 08:32:05.795 ERROR (qtp1756573246-34-null-1) [c: s: r: x:core_name
t:null-1] o.a.s.s.MultiThreadedSearcher raw read max=5922019 {code}
The max number is the total number of documents in the core.
I've tracked it down to this part of the code:
[https://github.com/apache/solr/blob/5bc7c1618e05b35bd0fa8471ae09329357a82036/solr/core/src/java/org/apache/solr/search/MultiThreadedSearcher.java#L86-L91]
I'm not entirely convinced that an ERROR level message is necessary here?
* The query seems to still function;
* Once the error condition is logged, the code seems to create a new doc set
and continues;
* The documentation doesn't suggest anything for how to avoid this? I'm not
sure why "needDocSet" is true here, and how it can be anything otherwise?
Surely an "info" or "warn" log message is more appropriate for these cases?
Unless it really is an error condition, but then the docs should be updated to
mention what could be done to avoid the error?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@solr.apache.org
For additional commands, e-mail: issues-h...@solr.apache.org
