Elasticsearch issue with some index/types, problem with reading table metadata

Hi,

I’ve got an error returning for many indexes in Elastic.
show as DATA_READ ERROR : Failure while attempting to read metadata for table

SQL : select * from

java.lang.NumberFormatException.forInputString():65

coming from writeDeclaredMap():317

seems liek a problem with analyzing the json, but I’m not sure what is it.

Can you share your indexes’ mappings? Post all the stack trace and more infos :slight_smile:

As I’ve the full stack trace, inside unattached network, I’ll give some parts only :
java.lang.NumberFormatException.forInputString():65

com.dremio.plugin.elastic.execution.ElasticSearchJsonReader.writeDeclaredMap(): 317

com.dremio.plugins.elastic.ElasticTableBullder.populate(): 183

I think it should be enough to understand the issue.

It might be due to the use of the ip data type, which has also representation in int format, maybe that is the issue.
About the mapping, its quite a big mapping, but it seems like the main difference from the point of view of types in the ip type, which seems like the issue here.

Note that elastic-search also has geo-spatial types there

Hi @ravenr,

Could you clarify by what you mean by the ip data type having a representation in int format? Dremio should support Elasticsearch IP types.

It would also be helpful to share the query profile as well as the verbose error message.

Thanks,
Rohan

It fails on preview of the index, so I cannot do any query on it.
It will probably happen on any created index with ip type in the mapping, so it should be easy to reproduce it.
Elastic-Search keeps also int representation of ip in order to allow efficient ip-range queries, I only wrote it cause it might explain the int format exception.

@ravenr I replied to your new post (Link) regarding how IP data types are supported through Dremio. Hopefully that should help you get up and running with Dremio.