[1/3] impala git commit: IMPALA-7425: Change incremental stats to pull from catalogd.

Repository: impala Updated Branches: refs/heads/master 3433d8fe1 -> 8dcf54aee IMPALA-7425: Change incremental stats to pull from catalogd. Currently, incremental stats can consume a substantial amount of metadata memory (per table, partition, column). This metadata is transmitted from

[2/3] impala git commit: Elaborating error message when incorrect port number is specified in impala-shell.

Elaborating error message when incorrect port number is specified in impala-shell. I recently helped debug an issue where impala-shell was being given the hiveserver2 port rather than the beeswax port. I've updated the error message a little bit to indicate that this may be the issue. Here is

[4/9] impala git commit: IMPALA-7490: fix uninitialized variables in load-data.py

IMPALA-7490: fix uninitialized variables in load-data.py Fixes use of an uninitialized variable in bin/load-data.py I found the following error message in a failed build, which is quite misleading: Traceback (most recent call last): File "bin/load-data.py", line 495, in if __name__ ==

[8/9] impala git commit: Fix generate_junitxml for python2.6.

Fix generate_junitxml for python2.6. With python 2.6, the syntax "{}".format(1) doesn't work: $docker run centos:6 python -c 'print "{}".format(1)' Traceback (most recent call last): File "", line 1, in ValueError: zero length field name in format generate_junitxml was using

[3/9] impala git commit: IMPALA-7508: Add Impala Python GDB module

IMPALA-7508: Add Impala Python GDB module This patch adds a new Impala Python GDB module, and implements a couple of covenience commands to make core dump analysis easier. The initial commands let us find queries and fragment instances currently executing in an impalad at the time the daemon

[5/9] impala git commit: IMPALA-7477: Batch-oriented query set construction

IMPALA-7477: Batch-oriented query set construction Rework the row-by-row construction of query result sets in PlanRootSink so that it materialises an output column at a time. Make some minor optimisations like preallocating output vectors and initialising strings more efficiently. My intent is

[1/9] impala git commit: IMPALA-6923: Remove create_database.py and perf_result_datastore.py

Repository: impala Updated Branches: refs/heads/master d4e281b73 -> 3433d8fe1 IMPALA-6923: Remove create_database.py and perf_result_datastore.py Description: The current upstream usage indicates that report_benchmark_results.py script is the only script in this folder that is used to

[9/9] impala git commit: Add bin/jenkins/finalize.sh

Add bin/jenkins/finalize.sh This provides a place for doing cleanup and error checking at the end of Jenkins builds. Currently, it only checks dmesg for OOMs. Testing: - https://jenkins.impala.io/job/ubuntu-16.04-from-scratch/3076/ Change-Id: If4d223b6e441f01a5edad9d6f9ca69c6de023461

[7/9] impala git commit: [diagnostics] Make --minidump_dir consistent with Impala's --minidump_path

[diagnostics] Make --minidump_dir consistent with Impala's --minidump_path Currently, the diagnostics script expects a full path to the actual directory to which process minidumps are written. This is however incosistent with Impala's configuration --minidump_path. Impala creates a subdirectory