You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Pavel Tupitsyn (Jira)" <ji...@apache.org> on 2021/12/17 18:23:00 UTC

[jira] [Created] (IGNITE-16152) Thin 3.0: Optimize server-side Tuple handling using known schema version

Pavel Tupitsyn created IGNITE-16152:
---------------------------------------

             Summary: Thin 3.0: Optimize server-side Tuple handling using known schema version
                 Key: IGNITE-16152
                 URL: https://issues.apache.org/jira/browse/IGNITE-16152
             Project: Ignite
          Issue Type: Improvement
          Components: thin client
    Affects Versions: 3.0.0-alpha3
            Reporter: Pavel Tupitsyn
            Assignee: Pavel Tupitsyn
             Fix For: 3.0.0-alpha4


Thin client protocol is schema-aware, but the known schema is thrown away on the server after deserialization, and Table API has to match it against schemas again.

Investigate if we can gain performance by preserving known schema using *SchemaAware* interface.

*Example: table.Upsert(Tuple t)*

# Client: match user tuple to schema, serialize with schema version, send to server
# Server: read schema version, read tuple data, create Tuple, throw away schema version, pass to Table API
# Server: match tuple to schema




--
This message was sent by Atlassian Jira
(v8.20.1#820001)