Sergey Soldatov created HDDS-16392:
--------------------------------------
Summary: parse Proto2Codec values from the CodecBuffer instead of
through an InputStream
Key: HDDS-16392
URL: https://issues.apache.org/jira/browse/HDDS-16392
Project: Apache Ozone
Issue Type: Improvement
Components: OM
Affects Versions: 2.3.0
Reporter: Sergey Soldatov
Assignee: Sergey Soldatov
Proto2Codec.fromCodecBuffer parsed the value through a ByteBufInputStream.
protobuf's CodedInputStream.newInstance(InputStream) allocates a 4 KB decoding
buffer per call, while the table values it decodes (OmKeyInfo, OmDirectoryInfo,
OmBucketInfo, ...) are typically a few hundred bytes. So every key an OM
listing or lookup parsed produced 4 KB of garbage on top of the value itself.
We might parse the buffer directly.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
