[
https://issues.apache.org/jira/browse/IGNITE-23114?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sergey Korotkov updated IGNITE-23114:
-------------------------------------
Description:
Currently the stream-related APIs for CLOBs and BLOBs are not supported in JDBC
drivers (both client and thin). Say the following ones:
* Blob.setBinaryStream
* PreparedStatement.setBlob(int parameterIndex, InputStream inputStream)
It makes the creation of BLOBs and CLOBs inefficient (memory wise) for large
data.
The whole data should be preloaded to RAM before the LOB creation.
After that several copies of the input memory blocks are created just to pass
it to prepared statement etc. It should be optimized to reduce the heap memory
requirements.
was:
Currently the stream-related APIs for CLOBs and BLOBs are not supported in JDBC
drivers (both client and thin). Say the following ones:
* Blob.setBinaryStream
* PreparedStatement.setBlob(int parameterIndex, InputStream inputStream)
It makes the creation of BLOBs and CLOBs inefficient memory wise.
The whole data should be preloaded to RAM before the LOB creation.
After that several copies of the input memory blocks are created just to pass
it to prepared statement etc. It should be optimized to reduce the heap memory
requirements.
> JDBC thin: support and optimize stream APIs for CLOBs and BLOBs
> ---------------------------------------------------------------
>
> Key: IGNITE-23114
> URL: https://issues.apache.org/jira/browse/IGNITE-23114
> Project: Ignite
> Issue Type: Task
> Reporter: Sergey Korotkov
> Priority: Major
> Labels: ise, jdbc
>
> Currently the stream-related APIs for CLOBs and BLOBs are not supported in
> JDBC drivers (both client and thin). Say the following ones:
> * Blob.setBinaryStream
> * PreparedStatement.setBlob(int parameterIndex, InputStream inputStream)
> It makes the creation of BLOBs and CLOBs inefficient (memory wise) for large
> data.
> The whole data should be preloaded to RAM before the LOB creation.
> After that several copies of the input memory blocks are created just to pass
> it to prepared statement etc. It should be optimized to reduce the heap
> memory requirements.
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
