Solr¶
The Wodby Solr catalog stack runs SolrCloud with a required ZooKeeper service. The current service composition and Solr
version are documented in wodby/stack-solr.
Authentication and endpoints¶
Wodby generates a password for SolrCloud's solr administrator account. Applications connected through a stack link
receive the internal hostname and generated credential defined by their service integration.
Solr exposes an HTTP endpoint on port 8983, so Wodby creates a technical route for its UI and API. Solr already uses
HTTP Basic authentication. Do not add Wodby route-level basic authentication to this route because an HTTP client cannot
send two independent Basic authorization headers.
Default collection¶
The Solr post-deployment initialization creates a collection named default when the cluster has no collections. It
uses Solr's _default configset.
_default is only a starting point. Applications that require a particular schema must install their own configset.
For Drupal, follow Connect Drupal to Solr so Search API Solr generates and uploads a compatible
configset.
Create another collection¶
You can create collections from the Solr Admin UI. You can also open a web terminal on the Solr app service and use the image action:
make create-collection collection=my_collection num_shards=1 config=_default \
-f /usr/local/bin/actions.mk
Replace my_collection, the shard count, and the configset for your application. The command uses the generated
SOLR_CLOUD_PASSWORD automatically.
Configsets¶
Keep custom configsets in a reproducible application or service workflow and upload them to ZooKeeper. Do not edit configsets bundled in the running image: those changes are tied to one container and disappear when it is replaced.
When a client integration can generate its own configset, prefer that workflow. After replacing a configset, follow the client's instructions for reloading the collection and reindexing data.
Use application backups for supported persistent service data and review volume settings before production use.