[ https://issues.apache.org/jira/browse/SOLR-11717?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16349408#comment-16349408 ]
Joel Bernstein edited comment on SOLR-11717 at 2/1/18 10:33 PM: ---------------------------------------------------------------- Let's also create a specific Solrj client for Streaming Expressions. You can use SolrStream but that was really meant to be a low level Stream not a user facing Stream. We could call it StreamClient or something like that. Sample syntax: {code:java} StreamClient client = new StreamClient(zkUrl, collection, params) ; client.open(); while(true){ Tuple t = client.read(); } client.close(); {code} was (Author: joel.bernstein): Let's also create a specific Solrj client for Streaming Expressions. You can use SolrStream but that was really meant to be a low level Stream not a user facing Stream. We could call it StreamClient or something like that. Sample syntax: {code:java} StreamClient client = new StreamClient(zkUrl, params) ; client.open(); while(true){ Tuple t = client.read(); } client.close(); {code} > Document Streaming Examples using SolrJ > --------------------------------------- > > Key: SOLR-11717 > URL: https://issues.apache.org/jira/browse/SOLR-11717 > Project: Solr > Issue Type: Improvement > Security Level: Public(Default Security Level. Issues are Public) > Components: documentation, streaming expressions > Reporter: Varun Thacker > Priority: Major > > Under http://lucene.apache.org/solr/guide/streaming-expressions.html it would > be nice if we had a page dedicated to how to use streaming with SolrJ > Topics we could cover > - How to correctly create a stream expression : > https://issues.apache.org/jira/browse/SOLR-11600?focusedCommentId=16261446&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-16261446 > - What is a SolrClientCache > - Which Stream objects should be a singleton and reused > - How many http connections / connections_per_host are we limited with by > default > - Example where the the Solr query node becomes the co-ordinator for the > streaming expressions ( {{TupleStream solrStream = new > SolrStream(solr_query_node, paramsLoc);}} > - How to use an "empty" Solr node to fire queries against. This node should > not have any collections. We could reduce the jetty thread pool size to rate > limit number of parallel queries potentially? -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org