[jira] [Updated] (IGNITE-22958) Improve performance of TupleMarshaller part 2

["Konstantin Orlov (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IGNITE-22958?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Konstantin Orlov updated IGNITE-22958:
--------------------------------------
    Description: This is follow up ticket to IGNITE-22941. In original ticket, only statistic collection has been addressed. Another significant part of tuple marshaling is access to the values, which involve quoting and normalization of names of the columns. We need to avoid normalization somehow, at least for internal calls when it's known that names in its normal form.  (was: This is follow up ticket to IGNITE-22941. In original ticket, only statistic collection has been addressed. Another significant part of tuple marshaling is access to the values, which involve quoting and normalization of names of the columns. It seems like we can safely omit names normalization at all, because schema contains columns with already normalized names.)

> Improve performance of TupleMarshaller part 2
> ---------------------------------------------
>
>                 Key: IGNITE-22958
>                 URL: https://issues.apache.org/jira/browse/IGNITE-22958
>             Project: Ignite
>          Issue Type: Improvement
>            Reporter: Konstantin Orlov
>            Assignee: Konstantin Orlov
>            Priority: Major
>              Labels: ignite-3
>
> This is follow up ticket to IGNITE-22941. In original ticket, only statistic collection has been addressed. Another significant part of tuple marshaling is access to the values, which involve quoting and normalization of names of the columns. We need to avoid normalization somehow, at least for internal calls when it's known that names in its normal form.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)