Saketa Chalamchala created HDDS-16230:
-----------------------------------------
Summary: Add cache-aware Table.multiGet for OM metadata batch reads
Key: HDDS-16230
URL: https://issues.apache.org/jira/browse/HDDS-16230
Project: Apache Ozone
Issue Type: Sub-task
Reporter: Saketa Chalamchala
Assignee: Saketa Chalamchala
Live OM tables use PARTIAL_CACHE / FULL_CACHE via TypedTable. A general
`Table.multiGet` would benefit callers that need batch reads with correct cache
behavior(Recon, quota repair, and other hot paths beyond snap-diff).
Extend [HDDS-16229|https://issues.apache.org/jira/browse/HDDS-16229] with
cache-integrated batch reads:
1. Table.multiGet(List<KEY>) — consult cache per key; issue one RocksDB
MultiGet for misses only
2. TypedTable.multiGet — partial/full cache semantics aligned with single-key
get()
3. DatanodeTable delegate; InMemoryTestTable implementation
4. RDBMetrics — batch get counters (calls and keys)
5. Identify and adopt 1–2 high-value call sites (separate small follow-ups if
needed)
Out of scope: CodecBuffer fast path (Tier 3), broad repo-wide migration of all
get() loops.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
