AB asked me about this offline and gave me reason to suspect that my 
*second* message about this (repeated below) bever made it to the list...

: : in downstream tests where I'm using LogListener and/or introspection hooks 
: : of some custom SearchComponents, I'm finding a lot of unexplained requests 
: : for 'q={!lucene}*:*' popping up in my tests.  At this point it's more of a 
: : nucance then an obvious indication of a "bug" -- and i'm not sure if it's 
: : a result of something that changed in the test-framework, or solrj, or 
: : solr-core -- but it certainly smells fishy.
: 
: I (sort of) tracked this down to changes in the PingRequestHandler.
: 
: My downstream code+tests are smart enough to ignore "ping" requests, but 
: they do so by inspecting the "path" of the request -- between 10.0 and the 
: 10.1 RC something changed (SOLR-18419 it appears) that has affected how 
: the ping requests internal processing is done so that the book keeping of 
: them as "ping" requests is now lost.
: 
: 
: Consider...
: 
: 1. docker run -p 8983:8983 --rm -it apache/solr:XXXX solr-demo
: 2. http://localhost:8983/solr/demo/admin/ping
: 
: ## With solr:10.0, that single ping request is logged as...
: 
: 2026-09-29 23:00:54.281 INFO  (qtp1745043985-37-null-12) [ x:demo t:null-12] 
o.a.s.c.S.Request path=/admin/ping params={} hits=45 status=0 QTime=3
: 2026-09-29 23:00:54.281 INFO  (qtp1745043985-37-null-12) [ x:demo t:null-12] 
o.a.s.c.S.Request path=/admin/ping params={} status=0 QTime=4
: 
: (first the internal request generated by the PingRequestHandler, then the 
: external /admin/ping request itself)
: 
: ## With apache/solr:10.1.0-SNAPSHOT that same single ping causes...
: 
: 2026-09-29 23:06:22.807 INFO  (qtp335580595-43-null-1) [ x:demo t:null-1] 
o.a.s.c.S.Request path=null 
params={{params(distrib=false),defaults(q={!lucene}*:*&echoParams=all)}} 
hits=45 status=0 QTime=21
: 2026-09-29 23:06:22.808 INFO  (qtp335580595-43-null-1) [ x:demo t:null-1] 
o.a.s.c.S.Request path=/admin/ping params={} status=0 QTime=41
: 
: ...i don't know if it's good or bad that that the internal request now 
: includes it's params, but it's certainly concerning that it has a null 
: "path" ... i'm not sure what about the change caused the path to start 
: being null, but it certainly smells like the kind of thing that might 
: break some code paths in Solr?


-Hoss
http://www.lucidworks.com/

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to