Ok, I could reproduce the issue quickly. Two things : - 1 : the MapreduceIndexerTool generates metadata about the file. These "fields" are presented to Solr https://www.cloudera.com/documentation/enterprise/5-6-x/topics/search_metadata.html So if you need these fields, you need to add them to the schema of the collection. - 2 : if you don't need these fields, then you have to delete them before presenting the row to Solr. For this, the function "sanitizeUnknownSolrFields" if the right way to go. But you missplaced it in your morhpline. It should be inside the command [] and after the readCSV function (of course before the loadSolr function). Something like this : commands [
{
readCSV {
...
}
}
{
sanitizeUnknowSolrFields {
...
}
}
{
loadSolr {
...
}
}
]
... View more
