On 7/17/2013 8:02 AM, Furkan KAMACI wrote: > At my indexing process to my SolrCloud(Solr 4.2.1) from Hadoop I got an > error. What is the reason, does it results with document loss for indexing? > > ERROR - 2013-07-17 16:30:01.453; org.apache.solr.common.SolrException; > java.lang.RuntimeException: [was class org.eclipse.jetty.io.EofException] > early EOF
Every time I've seen this particular exception, it has been caused by the client disconnecting before the request completes. If you are seeing it when doing index updates, then your client is behaving badly and not following the HTTP and/or TCP specifications, or it has been defined with a socket timeout that is not long enough for the update to complete. It's generally not a good idea to define a socket timeout for any kind of request other than a query. As for whether it is resulting in data loss, only you can answer that. Is the stacktrace preceded or followed in the log by an update request that has a list of ID values? If so, then the updates are probably working, but your client is not getting the response. Thanks, Shawn
