Alexander Behm commented on KUDU-1695:

Sounds like doing the row layout check on a per batch basis could be expensive 
for rows with few columns (relative to the savings from not just memcpying). 
I'd prefer to move that check out of the hot path and maybe declare the intent 
of the client to acquire ownership of the memory earlier, maybe when opening a 
Kudu scanner?

> C++ client support for zero copy scans
> --------------------------------------
>                 Key: KUDU-1695
>                 URL: https://issues.apache.org/jira/browse/KUDU-1695
>             Project: Kudu
>          Issue Type: New Feature
>          Components: client, perf
>    Affects Versions: 1.0.0
>            Reporter: Matthew Jacobs
>              Labels: impala
> Copying tuples from the Kudu client is expensive. Impala would like to be 
> able avoid this copy if the Kudu client can provide a mechanism to transfer 
> the memory ownership.

This message was sent by Atlassian JIRA

Reply via email to