Error log file - :ORDER/GROUP BY expression not found in targetlist

eli_revach

Hi , During data load my Error log file include many messages as the below (I like to mention that my load process is the only processing being executed on the system – the load is the single table with super projection only ) : :DCErrors time:424101378883039 node_name:v_bsm_node0002 session_id:MYDPHDB0079-31519:0x432 user_id:45035996273704962 user_name:dbadmin transaction_id:45035996280004912 statement_id:1 request_id:31 error_level:20 file_name:tlist.c line_number:202 function_name:get_sortgroupclause_tle message:ORDER/GROUP BY expression not found in targetlist error_code:16779814 vertica_code:4300 detail: hint: log_message:ORDER/GROUP BY expression not found in targetlist log_detail: log_hint: cursor_position:0

[Deleted User]

Hi Eli, Hm... I don't know the cause of this error. I assume that you are not also seeing the error as output from vsql, or reported as a JDBC/ODBC error? If not, then the error is most likely harmless. But it should not be appearing in the dc_errors table. (Incidentally, I would recommend reading that table with "select * from dc_errors", rather than looking at the log file directly. Vertica buffers DC-table records in memory; this minimizes the performance impact of the extra logging, but it also means that the underlying data files in question are often missing the latest records. "Also that's our abstraction barrier; no peeking at the implementation!" :-) ) I do have a few theories about what might be going on, though all are kinda fuzzy. Do you have an example sequence of statements that consistently causes this error to be emitted? Preferably simple. One hypothesis (not the only one) involves the first COPY into a new table. Most of my ideas have at least some dependency on the exact table definition / DDL. Adam

ilselieve

I got this error when I was using PostgreSQL to query an Amazon Redshift database. To fix it, I had to change the RANK function to this Redshift-specific function:

SUM (1) OVER ( PARTITION BY field1 ORDER BY field2 ROWS UNBOUNDED PRECEDING)