Hello Quanlong Huang, Michael Smith, Impala Public Jenkins,
I'd like you to reexamine a change. Please visit
http://gerrit.cloudera.org:8080/24068
to look at the new patch set (#2).
Change subject: IMPALA-13122 addendum: Fix host statistics logging for erasure
coded files
......................................................................
IMPALA-13122 addendum: Fix host statistics logging for erasure coded files
When erasure coding is enabled, disk IDs are unavailable for
EC blocks. The previous implementation only tracked hosts via host:disk
pairs, requiring valid disk IDs. This caused host statistics to be missing
from logs in EC environments.
Fixed by tracking host indices separately from host:disk pairs:
- Added uniqueHostIndices set to FileMetadataStats
- Track all host indices regardless of disk ID availability
- Host:disk pairs still tracked only when disk IDs are valid (>= 0)
- Updated getNumUniqueHosts() to use uniqueHostIndices directly
With this fix:
- Traditional replication: Both hosts and host:disk pairs are logged
- Erasure coding: Hosts are logged, host:disk pairs may be 0 or omitted
Testing:
- All tests pass with and without erasure coding
Change-Id: Ie6f5b70fa9c46dd3f34287f030553360da6b20c6
---
M fe/src/main/java/org/apache/impala/catalog/FeFsTable.java
M fe/src/test/java/org/apache/impala/catalog/FileMetadataLoaderTest.java
M tests/common/environ.py
M tests/custom_cluster/test_file_metadata_stats.py
4 files changed, 23 insertions(+), 14 deletions(-)
git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/68/24068/2
--
To view, visit http://gerrit.cloudera.org:8080/24068
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: newpatchset
Gerrit-Change-Id: Ie6f5b70fa9c46dd3f34287f030553360da6b20c6
Gerrit-Change-Number: 24068
Gerrit-PatchSet: 2
Gerrit-Owner: Arnab Karmakar <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Michael Smith <[email protected]>
Gerrit-Reviewer: Quanlong Huang <[email protected]>
