Querying data in a distributed storage system

A data node of a distributed storage system, comprising:

• a logic machine;
• local storage;
• and
• memory comprising instructions executable by the logic machine to
• receive, from a front-end node, a query regarding data in a block of an object, the block being stored in the local storage, the query comprising one or more predicates;
• execute the query on the data in the block to identify one or more records to return to the front-end node, and to identify an incomplete record that extends to an adjacent block of the object;
• and
• send, to the front-end node, results for the query and the incomplete record.

A front-end node of a distributed storage system, comprising:

• a logic machine;
• and
• a storage machine comprising instructions executable by the logic machine to
• receive a query regarding data stored in an object, the query comprising one or more predicates;
• obtain information identifying a plurality of data nodes at which blocks of the object are stored;
• send, to a first data node at which a first block of the object is stored, a query regarding data in the first block based upon the one or more predicates;
• receive, from the first data node, first results for the query regarding the data in the first block, the first results comprising an incomplete record;
• send, to a second data node at which a second block of the object is stored, a query regarding data in the second block based upon the one or more predicates, the query comprising the incomplete record, the second block being stored on a different physical device than the first block;
• receive, from the second data node, results for the query regarding the data in the second block;
• and
• output the first results and the second results for the queries regarding the data in the first block and the data in the second block.