On 10/29/2014 3:09 AM, Diego Marconato wrote: > with Solr 3.3.0 when indexing I get the following errors (sometimes): > > org.apache.solr.common.SolrException log > org.apache.solr.common.SolrException: Invalid chunk header > Caused by: com.ctc.wstx.exc.WstxIOException: Invalid chunk header > Caused by: java.io.IOException: Invalid chunk header > at > org.apache.coyote.http11.filters.ChunkedInputFilter.throwIOException(ChunkedInputFilter.java:610) > > > Any idea of a workaround to solve the problem ..... I can act in some way > in solrconfig.xml?
The class org.apache.coyote.http11.filters.ChunkedInputFilter is a tomcat class. Tomcat seems to be complaining about the HTTP connection from the client. HTTP chunking (a feature of http 1.1, related to keepalive if I'm not mistaken) seems to be having problems. Solr itself is not involved with this exception. This is most likely a client issue, a bug in tomcat, or a configuration issue in tomcat. The client issue could be an issue with a proxy, too. The haproxy load balancer (which I use with Solr) usually requires the http-pretend-keepalive option when talking to a servlet container like jetty or tomcat. http://cbonte.github.io/haproxy-dconv/configuration-1.5.html#4-option%20http-pretend-keepalive Thanks, Shawn
