There are various reasons - mainly we want more control over memory usage
and accounting than shared_ptr allows.
Generally we avoid shared_ptr in Impala since it makes it harder to reason
about when resources are released. E.g. we typically want to know/control
exactly when memory is freed up.
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4049: fix empty batch handling NLJ build side
..
Patch Set 2:
(1 comment)
It seems like it would be very difficult to trigger without a subplan
-ASF
Gerrit-Branch: master
Gerrit-Owner: Tim Armstrong <tarmstr...@cloudera.com>
Gerrit-Reviewer: Alex Behm <alex.b...@cloudera.com>
Gerrit-Reviewer: Matthew Jacobs <m...@cloudera.com>
Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com>
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3308: Get expr-test passing on PPC64LE
..
Patch Set 1:
* from what I can see only the ExprValue() and ExprValue(std::string&)
constructors are
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3308: Get expr-test passing on PPC64LE
..
Patch Set 1:
Actually I think we should go further to prevent such bugs in future. Currently
it's still
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3308: Get expr-test passing on PPC64LE
..
Patch Set 1:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4186/1/be/src/exprs/literal.cc
File be/src/exprs
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4051: Avoid unnecessary copy of RowDescriptor into
RowBatch
..
Patch Set 1:
Created IMPALA-4051. It sometimes feels like overkill to create JIRAs
Looks good to me - thank you for putting this together. I see you've
already identified minidump2core as a todo - that would also be helpful
since sometimes that lets you inspect local variables and arguments, which
I don't think you can do with the stackwalk tool.
On Wed, Aug 31, 2016 at 3:42
Tim Armstrong has uploaded a new patch set (#2).
Change subject: IMPALA-4051: Avoid unnecessary copy of RowDescriptor into
RowBatch
..
IMPALA-4051: Avoid unnecessary copy of RowDescriptor into RowBatch
This commit changes
Tim Armstrong has uploaded a new patch set (#2).
Change subject: IMPALA-4049: fix empty batch handling NLJ build side
..
IMPALA-4049: fix empty batch handling NLJ build side
Memory from the build side of a nested loop join
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4182
Change subject: IMPALA-4049: fix empty batch handling NLJ build side
..
IMPALA-4049: fix empty batch handling NLJ build side
Memory
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4181
Change subject: Avoid unnecessary copy of RowDescriptor into RowBatch
..
Avoid unnecessary copy of RowDescriptor into RowBatch
Tim Armstrong has posted comments on this change.
Change subject: Revert "Use impala-python when building shell tarball"
..
Patch Set 1: Code-Review+2
--
To view, visit http://gerrit.cloudera.org:8080/4176
To u
Tim Armstrong has uploaded a new patch set (#4).
Change subject: IMPALA-4037,IMPALA-4038: fix locking during query cancellation
..
IMPALA-4037,IMPALA-4038: fix locking during query cancellation
* Refactor the child query
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-2831: Bound the number of scanner threads per scan node.
..
Patch Set 1:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4174/1/be/src/exec/hdfs-scan
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-2831: Bound the number of scanner threads per scan node.
..
Patch Set 1:
(2 comments)
http://gerrit.cloudera.org:8080/#/c/4174/1/be/src/exec/hdfs-scan
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4019: initialize member variables in HdfsTableSink
..
Patch Set 1: -Code-Review
--
To view, visit http://gerrit.cloudera.org:8080/4171
To unsubscribe
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4019: initialize member variables in HdfsTableSink
..
Patch Set 1: Code-Review-2
I started a GVO. Marking -2 so it won't accidentally get merged
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4171
Change subject: IMPALA-4019: initialize member variables in HdfsTableSink
..
IMPALA-4019: initialize member variables in HdfsTableSink
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3905: Add single-threaded scan node.
..
Patch Set 3:
(10 comments)
http://gerrit.cloudera.org:8080/#/c/4113/3/be/src/exec/hdfs-avro-scanner.cc
File
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3808: Add incubating DISCLAIMER from the Incubator
Branding Guide
..
Patch Set 1: Code-Review+1
--
To view, visit http://gerrit.cloudera.org:8080/4160
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4027:Memory leak with ExprCtxs not free
..
Patch Set 2: Code-Review+2
Rebased, carry +2
--
To view, visit http://gerrit.cloudera.org:8080/4132
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3671: Add query option to limit scratch space usage
..
Patch Set 5:
(20 comments)
http://gerrit.cloudera.org:8080/#/c/3938/4/be/src/runtime/buffered
Tim Armstrong has uploaded a new patch set (#5).
Change subject: IMPALA-3671: Add query option to limit scratch space usage
..
IMPALA-3671: Add query option to limit scratch space usage
Currently we can only disable spilling
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4027:Memory leak with ExprCtxs not free
..
Patch Set 1: Code-Review+1
The change looks good to me, nice catch.
Have you submitted a license agreement
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3201: reservation implementation for new buffer pool
..
Patch Set 7:
(1 comment)
Added reservation counters and reporting, plus simplified the linkage
Tim Armstrong has uploaded a new patch set (#8).
Change subject: IMPALA-3201: reservation implementation for new buffer pool
..
IMPALA-3201: reservation implementation for new buffer pool
This patch implements the reservation
Tim Armstrong has uploaded a new patch set (#7).
Change subject: IMPALA-3201: reservation implementation for new buffer pool
..
IMPALA-3201: reservation implementation for new buffer pool
This patch implements the reservation
A blog post related to some work going on in the parquet-cpp project to
speed up dictionary decoding. It's still a very rough prototype. Might be
worth keeping an eye on.
http://lemire.me/blog/2016/08/25/faster-dictionary-decoding-with-simd-instructions/
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3905: Add single-threaded scan node.
..
Patch Set 1:
(6 comments)
A few more comments. I think I understand the core of this a bit better - will
wait
Tim Armstrong has posted comments on this change.
Change subject: System Database (preview for frontend)
..
Patch Set 13:
(17 comments)
I took another look over this, looks like we're making progress. I know you
have other
+1 (Binding)
On Thu, Aug 25, 2016 at 3:18 PM, Matthew Jacobs wrote:
> +1 (binding)
>
> On Thu, Aug 25, 2016 at 3:01 PM, Marcel Kornacker
> wrote:
> > +1 (binding)
> >
> > On Thu, Aug 25, 2016 at 1:48 PM, Jim Apple wrote:
> >> Oh,
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4006: dangerous rm -rf statements in scripts
..
Patch Set 6:
We already have set -euo pipefail at the top of all of the scripts in the repo,
which turns
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-1659: Netezza compatibility functions: metadata
..
Patch Set 6: Code-Review+1
--
To view, visit http://gerrit.cloudera.org:8080/4063
To unsubscribe
Tim Armstrong has uploaded a new patch set (#10).
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
The main outcome of this patch is to split out
Tim Armstrong has posted comments on this change.
Change subject: Stricter clang-format: set DerivePointerAlignment to false.
..
Patch Set 1: Code-Review+2
--
To view, visit http://gerrit.cloudera.org:8080/4127
To unsubscribe
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
Patch Set 9:
(17 comments)
http://gerrit.cloudera.org:8080/#/c/3873/9/be/src/exec/analytic-eval
Tim Armstrong has abandoned this change.
Change subject: IMPALA-2809: Improve ByteSwap with builtin function or SSSE3 or
AVX2.
..
Abandoned
Moved to https://gerrit.cloudera.org/#/c/4124/1
--
To view, visit http
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-1659: Netezza compatibility functions: metadata
..
Patch Set 4: Code-Review+1
(2 comments)
http://gerrit.cloudera.org:8080/#/c/4063/4/be/src/exprs/expr
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3905: Add single-threaded scan node.
..
Patch Set 1:
(22 comments)
I did an initial pass. I still need to think through some of the details but I
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3979: Fix dynamic linking for Impala
..
Patch Set 3: Code-Review+2
--
To view, visit http://gerrit.cloudera.org:8080/4108
To unsubscribe, visit http
Tim Armstrong has submitted this change and it was merged.
Change subject: IMPALA-3979: Fix dynamic linking for Impala
..
IMPALA-3979: Fix dynamic linking for Impala
When building Impala, the user can choose between two
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3979: Fix dynamic linking for Impala
..
Patch Set 3:
I'll go ahead and merge this. I verified it builds on all supported OSes
--
To view, visit http
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3979: Fix dynamic linking for Impala
..
Patch Set 2: Code-Review+2
Carry +2
--
To view, visit http://gerrit.cloudera.org:8080/4108
To unsubscribe
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4117
Change subject: Remove unused MemTracker debug logging
..
Remove unused MemTracker debug logging
The MemTracker code included debug
The fix works for me too, thanks Martin!
I gave the patch a +1 but will upgrade to +2 if it seems everyone is happy
with it.
Cheers,
Tim
On Wed, Aug 24, 2016 at 8:15 AM, Yonghyun Hwang
wrote:
> Hello Martin,
>
> Your fix works great! :) As for the build failure, what
Tim Armstrong has posted comments on this change.
Change subject: Fix dynamic linking for Impala
..
Patch Set 1:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4108/1//COMMIT_MSG
Commit Message:
Line 7: Fix dynamic linking
Tim Armstrong has posted comments on this change.
Change subject: Fix dynamic linking for Impala
..
Patch Set 1: Code-Review+1
The fix works for me and this is cleaner. As far as I can tell the webserver
was only a separate
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3832: test invalid data handling in lzo text scanner
..
Patch Set 3:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4096/3/tests/query_test
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3832: test invalid data handling in lzo text scanner
..
Patch Set 4: Code-Review+2
Carry +2
--
To view, visit http://gerrit.cloudera.org:8080/4096
: Ib707014c1fcfb80cb8076f644fc2b62a5ae758d7
Gerrit-PatchSet: 4
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Tim Armstrong <tarmstr...@cloudera.com>
Gerrit-Reviewer: Dan Hecht <dhe...@cloudera.com>
Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com>
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3945: Forbid create text table with nonsensical
delimiter combinations.
..
Patch Set 3:
Can we abandon this for now if we don't have any intent to merge
Tim Armstrong has abandoned this change.
Change subject: IMPALA-2700: ASCII NUL characters are doubled on insert into
text tables
..
Abandoned
This was already merged to master: https://gerrit.cloudera.org/#/c/3876
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3832: test invalid data handling in lzo text scanner
..
Patch Set 2:
(2 comments)
http://gerrit.cloudera.org:8080/#/c/4096/2/tests/query_test
Tim Armstrong has uploaded a new patch set (#3).
Change subject: IMPALA-3832: test invalid data handling in lzo text scanner
..
IMPALA-3832: test invalid data handling in lzo text scanner
This adds the lzo text scanner
Tim Armstrong has posted comments on this change.
Change subject: Add .clang-format for Impala's C++ style
..
Patch Set 6:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/3886/6/.clang-format
File .clang-format:
Line 1
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4096
Change subject: IMPALA-3832: test invalid data handling in lzo text scanner
..
IMPALA-3832: test invalid data handling in lzo text
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4006: dangerous rm -rf statements in scripts
..
Patch Set 3: Code-Review+1
Will wait for Michael's +1 too.
--
To view, visit http://gerrit.cloudera.org
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-1659: Netezza compatibility functions: metadata
..
Patch Set 2:
(4 comments)
http://gerrit.cloudera.org:8080/#/c/4063/2/be/src/exprs/expr-test.cc
File
Tim Armstrong has posted comments on this change.
Change subject: Infomation_schema (preview for frontend)
..
Patch Set 6:
(18 comments)
Did a pass over it and made some general comments. I think we should do some
cleanup
Tim Armstrong has posted comments on this change.
Change subject: Infomation_schema (preview for frontend)
..
Patch Set 6:
Could you run clang-format on the backend to clean up the formatting?
If you get .clang_format from
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3201: reservation implementation for new buffer pool
..
Patch Set 6:
Updated to use the UsedReservation/UnusedReservation terminology I discussed
Tim Armstrong has uploaded a new patch set (#6).
Change subject: IMPALA-3201: reservation implementation for new buffer pool
..
IMPALA-3201: reservation implementation for new buffer pool
This patch implements the reservation
Tim Armstrong has uploaded a new patch set (#9).
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
The main outcome of this patch is to split out
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
Patch Set 8:
Rebased this onto master MemLimitExceeded() changes. Had to make a few
straightforward
Tim Armstrong has uploaded a new patch set (#8).
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
The main outcome of this patch is to split out
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4006: impala-config.sh contains dangerous rm -rf
statements
..
Patch Set 2:
(2 comments)
Mostly looks good (aside from LOG_DIR that needs to be added
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-4006 impala-config.sh contains dangerous rm -rf
statements
..
Patch Set 1:
(13 comments)
This seems like a good practice, but I'm not clear on how you
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3981: Fix crash when accessing statestored / catalogd
/memz page
..
Patch Set 10: Code-Review+2
+2 after rebase
--
To view, visit http
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-(3895,3859): Don't log file data on parse errors
..
Patch Set 3: Code-Review+1
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4020/3/tests/common
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3943: Adhere to abort_on_error in ProcessFooter().
..
Patch Set 7:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/3862/7/be/src/exec/hdfs-parquet
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3662: Don't double allocate tuples buffer in parquet
scanner
..
Patch Set 2: Code-Review+1
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4064/1
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3662: Don't double allocate tuples' buffer in parquet
scanner
..
Patch Set 1:
(2 comments)
http://gerrit.cloudera.org:8080/#/c/4064/1/be/src/exec/hdfs
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-2033: Netezza compatibility functions quote_ident
..
Patch Set 4:
Hi Shirish, have you had a chance to look at the comments?
--
To view, visit http
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-1659: Netezza compatibility functions: metadata
..
Patch Set 1:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4063/1/be/src/exprs/utility-functions
hat is a bug fix.
>
> On Fri, Aug 19, 2016 at 9:01 AM, Tim Armstrong <tarmstr...@cloudera.com>
> wrote:
> > How do you plan to choose which commits to cherry-pick? Should we let you
> > know if we think a patch should/shouldn't be part of the release?
> >
> &
How do you plan to choose which commits to cherry-pick? Should we let you
know if we think a patch should/shouldn't be part of the release?
On Fri, Aug 19, 2016 at 8:44 AM, Tom White wrote:
> Thanks for volunteering to do the release Jim! The plan looks fine to me.
>
> Tom
>
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3943: Adhere to abort_on_error in ProcessFooter().
..
Patch Set 7:
(1 comment)
Not this patch, but we should consider changing HdfsScanNode so
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3342: Runtime profile TotalCpuTime should eliminate wait
times
..
Patch Set 1:
Unless I'm missing something, TotalCpuTime at the fragment level isn't
Tim Armstrong has posted comments on this change.
Change subject: Add .clang-format for Impala's C++ style
..
Patch Set 5: Code-Review+1
--
To view, visit http://gerrit.cloudera.org:8080/3886
To unsubscribe, visit http
Tim Armstrong has posted comments on this change.
Change subject: Add functional and targeted perf tests for joins with empty
builds
..
Patch Set 3: Code-Review+2
carry +2
--
To view, visit http://gerrit.cloudera.org:8080
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3090: always log memory limit errors
..
Patch Set 3: Code-Review+2
Rebase
--
To view, visit http://gerrit.cloudera.org:8080/4049
To unsubscribe, visit
Sounds reasonable to me.
On Thu, Aug 18, 2016 at 4:04 PM, Jim Apple wrote:
> http://gerrit.cloudera.org:8080/4053
>
> I am proposing we change the bylaws so that a branch can be created or
> deleted by lazy consensus of the PMC: "Lazy consensus requires no -1
> votes
Tim Armstrong has posted comments on this change.
Change subject: Update bylaws: Lazy Consensus for branch creation and deletion.
..
Patch Set 1: Code-Review+1
--
To view, visit http://gerrit.cloudera.org:8080/4053
Tim Armstrong has posted comments on this change.
Change subject: Add functional and targeted perf tests for joins with empty
builds
..
Patch Set 1:
(3 comments)
http://gerrit.cloudera.org:8080/#/c/4051/1/testdata/workloads
Tim Armstrong has uploaded a new patch set (#2).
Change subject: Add functional and targeted perf tests for joins with empty
builds
..
Add functional and targeted perf tests for joins with empty builds
I wrote these tests
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4051
Change subject: Add functional and targeted perf tests for joins with empty
builds
..
Add functional and targeted perf tests for joins
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-(3895,3859): Don't log file data on parse errors
..
Patch Set 2:
(1 comment)
http://gerrit.cloudera.org:8080/#/c/4020/2/testdata/workloads/functional
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3090: always log memory limit errors
..
Patch Set 1:
(2 comments)
http://gerrit.cloudera.org:8080/#/c/4049/1/be/src/runtime/mem-tracker.h
File be/src
Tim Armstrong has uploaded a new patch set (#2).
Change subject: IMPALA-3090: always log memory limit errors
..
IMPALA-3090: always log memory limit errors
Consistently log memory limit errors so that the error message contains
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-(3895,3859): Don't log file data on parse errors
..
Patch Set 2:
(4 comments)
It's kind of unfortunate to lose the diagnostic info if people are running
Tim Armstrong has uploaded a new change for review.
http://gerrit.cloudera.org:8080/4049
Change subject: IMPALA-3090: always log memory limit errors
..
IMPALA-3090: always log memory limit errors
Consistently log memory
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
Patch Set 7:
I figured out what was going on with the profiles - the builder profile was
being added
Tim Armstrong has uploaded a new patch set (#7).
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
The main outcome of this patch is to split out
Tim Armstrong has posted comments on this change.
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
Patch Set 5:
(17 comments)
Thanks for looking over the changes. I know it's a big patchset. Let me
Tim Armstrong has uploaded a new patch set (#6).
Change subject: IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
..
IMPALA-3567 Part 2, IMPALA-3899: factor out PHJ builder
The main outcome of this patch is to split out
Tim Armstrong has posted comments on this change.
Change subject: Add .clang-format for Impala's C++ style
..
Patch Set 3:
Shouldn't the ternary operator be affected by BreakBeforeTernaryOperators
though? Or is that something
but I also think we
> > should be willing to update it sometimes. I think when we do update
> > it, we don't need to do a bulk reformat.
> >
> > On Tue, Aug 16, 2016 at 9:06 AM, Tim Armstrong <tarmstr...@cloudera.com>
> wrote:
> >> +1 for automating thi
Tim Armstrong has posted comments on this change.
Change subject: Add .clang-format for Impala's C++ style
..
Patch Set 3: Code-Review+1
Sounds ok to me then
--
To view, visit http://gerrit.cloudera.org:8080/3886
Tim Armstrong has posted comments on this change.
Change subject: Add .clang-format for Impala's C++ style
..
Patch Set 2:
I ran this on partitioned-aggregation-node.cc and mostly the changes were
improvements (i.e
301 - 400 of 1592 matches
Mail list logo