dsmiley commented on code in PR #3904:
URL: https://github.com/apache/solr/pull/3904#discussion_r2659817051
##########
solr/solr-ref-guide/modules/deployment-guide/pages/solrcloud-recoveries-and-write-tolerance.adoc:
##########
@@ -59,3 +59,14 @@ For instance, a client application may want to keep a log of
which update reques
NOTE: In previous version of Solr, the `min_rf` parameter had to be specified
to ask Solr for the achieved replication factor.
Now it is always included in the response.
+
+== Solr Core Lifecycle Inconsistency
Review Comment:
I appreciate you want to document this, and that this page isn't a
terrible/obviously-wrong choice, but IMO it further degrades a very important
page that should be more focused on how key aspects of SolrCloud works.
Perhaps what we should do is log a warning with this suggestion that will
change. Maybe also reference the JIRA issue where an interested user can get
more background / advice. After all, it's an obscure matter. WDYT?
##########
solr/core/src/java/org/apache/solr/core/CoreContainer.java:
##########
@@ -1511,6 +1512,16 @@ public SolrCore create(
preExistingZkEntry =
getZkController().checkIfCoreNodeNameAlreadyExists(cd);
}
+ final boolean deleteUnknownCores =
+
EnvUtils.getPropertyAsBool("solr.cloud.delete.unknown.cores.enabled", false);
+ if (deleteUnknownCores && Files.exists(cd.getInstanceDir())) {
Review Comment:
Shouldn't this logic be adjacent to the logic that fails in
`CoreProperitesLocator.exists`? (I assume that's the spot). Doing so would be
better organized and probably lead to combining in a more clear way with
coordinated logging and existence check. I forget if this was discussed already
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
