It's under documented, under tested and SolrCloud doesn't like configurable end points like this. It's also documented as non operational. +1 on removing.
- Mark On Wed, May 23, 2018 at 4:56 PM Gus Heck <gus.h...@gmail.com> wrote: > I was just trying to understand the role of ShardHandler, and tracing back > things to understand when it's called (or not) with respect to updates. I > found myself reading the init() method of HttpSolrCall (because a lot of > logic around request types happens on init() ) and the very first thing in > the method is an if block regarding some sort of alternate path > substitution involving "managementPath"... This is apparently a valid > config item in solr.xml, so trying to figure out if I cared about this or > not I went to find out what it was... there was a confusing comment in > code, and a example of it being set in the file solr-50--all.xml, but that > appears to just test that setting it in the xml causes it to appear in the > config java object.... and then I finally found the documentation for it in > the ref guide: > > managementPath > > Currently non-operational. > > (https://lucene.apache.org/solr/guide/7_3/format-of-solr-xml.html) > > In other words I just spent 15-20 min tracking down something non-useful > that apparently was originally added in SOLR-695 (which says it was > committed to encourage review, but nobody other than the author commented > on it thereafter) > > This looks a lot like dead, or at least questionable code... perhaps it > should be removed so that folks don't have to spend time figuring out that > it is non-operational any time they read the code? > > Or is it operational/useful but undocumented? > > -Gus > > -- > http://www.the111shift.com > -- - Mark about.me/markrmiller