[jira] [Commented] (IMPALA-15117) Optimize getFirstLevelAcidDirPath

["Sai Hemanth Gantasala (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IMPALA-15117?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18094117#comment-18094117 ] 

Sai Hemanth Gantasala commented on IMPALA-15117:
------------------------------------------------

Impala passes the file path; Hive passes p.getParent().

 So Impala does one extra isDirectory per file. Hive actually loads ACID layout for reads/compaction, it already does a single listing pass:

getAcidState() → listStatus / tryListLocatedHdfsStatus, or getHdfsDirSnapshots().
isAcid() → one listStatus(partitionDir) and local filtering.
getHdfsDirSnapshots() → BFS via listStatusIterator, not per-subdir isDirectory.
So the O(N) RPC → 1 listStatus idea is largely already how Hive works for partition ACID discovery.

In summary, this optimization is not applicable for Hive. Thanks [~MikaelSmith] for asking to verify this in Hive code path.

> Optimize getFirstLevelAcidDirPath
> ---------------------------------
>
>                 Key: IMPALA-15117
>                 URL: https://issues.apache.org/jira/browse/IMPALA-15117
>             Project: IMPALA
>          Issue Type: Improvement
>          Components: Frontend
>    Affects Versions: Impala 4.5.0
>            Reporter: Michael Smith
>            Assignee: Jiyoung Yoo
>            Priority: Major
>              Labels: newbie, performance
>
> This could also apply to Hive.
> When loading ACID table metadata, {{AcidUtils.getFirstLevelAcidDirPath}} checks whether each candidate ACID subdirectory name (e.g. {{delta_*}}, {{base_*}}, {{delete_delta_*}}) exists under a partition directory by calling {{FileSystem.isDirectory(path)}} for each candidate in a loop. On distributed filesystems each call is a separate RPC.
> Observed via async-profiler on a production Impala catalogd connected to Apache Ozone (OFS): 258 samples (94% blocked on RPC) in the stack {{AcidUtils.getFirstLevelAcidDirPath → FileSystem.isDirectory → OzoneBucket.getFileStatus → OM gRPC call}}. For a partition with N candidate paths, N separate {{GetFileStatus}} RPCs are issued where 1 {{listStatus}} would suffice.
> We could replace the N {{isDirectory(candidatePath)}} calls with a single {{listStatus(partitionDir)}} call and resolve membership locally:
> # Call {{fs.listStatus(partitionDir)}} once to get all immediate children with their type flags.
> # Build a local {{Set<String>}} of child directory names from the result.
> # Filter candidate ACID names against the set — no additional RPCs.
> This reduces the cost from O(N) RPCs to 1 regardless of filesystem implementation. On object-store-backed filesystems (OFS, S3, ABFS) where each {{isDirectory}} is a synchronous RPC, the reduction is especially significant.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-all-unsubscribe@impala.apache.org
For additional commands, e-mail: issues-all-help@impala.apache.org