Hello Impala Public Jenkins,
I'd like you to reexamine a change. Please visit
http://gerrit.cloudera.org:8080/21257
to look at the new patch set (#4).
Change subject: IMPALA-12980: Translate CpuAsk into admission control slots
......................................................................
IMPALA-12980: Translate CpuAsk into admission control slots
Impala has a concept of "admission control slots" - the amount of
parallelism that should be allowed on an Impala daemon. This defaults to
the number of processors and can be overridden with
-–admission_control_slots flag.
Admission control slot accounting is described in IMPALA-8998. It
compute 'slots_to_use' for each backend based on the max number of
instances of any fragment on that backend. This is simplistic, because
multiple non-blocking fragments with the same number of instance counts,
say 4 non-blocking fragments each with 12 instances, only request the
max instance (12) admission slots rather than sum of it (48), making
each of 12 cores oversubscribed by 4x.
When COMPUTE_PROCESSING_COST is enabled, Planner will generate a CpuAsk
number that represents the cpu requirement of that query over a
particular executor group set. This number is an estimation of the
largest number of query fragment instances that can run in parallel
without waiting, given by the blocking operator analysis. Therefore, the
fragment trace that sums into that CpuAsk number can be translated into
'slots_to_use' as well, which will be a closer resemblance of maximum
parallel execution of fragment instances.
This patch adds a new query option called SLOT_COUNT_STRATEGY to control
which admission control slot accounting to use. There are two possible
values:
- LARGEST_FRAGMENT, which is the original algorithm from IMPALA-8998.
This is still the default value for the SLOT_COUNT_STRATEGY option.
- PLANNER_CPU_ASK, which will follow fragment trace that contributes
towards CpuAsk number.
To do the PLANNER_CPU_ASK strategy, the Planner will mark fragments that
contribute to CpuAsk as dominant fragments. It also passes
max_slot_per_executor information that it knows about the executor group
set to the scheduler.
AvgAdmissionSlotsPerExecutor counter is added to describe what Planner
think the average 'slots_to_use' per backend will be. Actual
'slots_to_use' in each backend may differ than
AvgAdmissionSlotsPerExecutor, depending on what scheduled on that
backend.
Testing:
- Update test_executors.py with AvgAdmissionSlotsPerExecutor assertion.
- Pass test_tpcds_queries.py::TestTpcdsQueryWithProcessingCost.
Change-Id: I338ca96555bfe8d07afce0320b3688a0861663f2
---
M be/src/scheduling/admission-controller.cc
M be/src/scheduling/scheduler.cc
M be/src/service/query-options.cc
M be/src/service/query-options.h
M common/thrift/ImpalaService.thrift
M common/thrift/Planner.thrift
M common/thrift/Query.thrift
M fe/src/main/java/org/apache/impala/common/Id.java
M fe/src/main/java/org/apache/impala/planner/CoreCount.java
M fe/src/main/java/org/apache/impala/planner/CostingSegment.java
M fe/src/main/java/org/apache/impala/planner/PlanFragment.java
M fe/src/main/java/org/apache/impala/planner/Planner.java
M fe/src/main/java/org/apache/impala/service/Frontend.java
M fe/src/test/java/org/apache/impala/planner/PlannerTestBase.java
M tests/custom_cluster/test_executor_groups.py
M tests/query_test/test_tpcds_queries.py
16 files changed, 312 insertions(+), 82 deletions(-)
git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/57/21257/4
--
To view, visit http://gerrit.cloudera.org:8080/21257
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: newpatchset
Gerrit-Change-Id: I338ca96555bfe8d07afce0320b3688a0861663f2
Gerrit-Change-Number: 21257
Gerrit-PatchSet: 4
Gerrit-Owner: Riza Suminto <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Riza Suminto <[email protected]>
