-
Notifications
You must be signed in to change notification settings - Fork 5.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add release notes for 0.241 #15143
Add release notes for 0.241 #15143
Conversation
4e20b72
to
5c87163
Compare
@kewang1024 @fgwang7w @beinan @zhenxiao I make changes to your commit messages. Could you please take a look to see if they're accurate? |
|
||
Hive Changes | ||
____________ | ||
* Fix several memory counting bugs in ``OrcRecordReader`` and ``StreamReader``'s. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
counting -> accounting reads better?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What is the trailing 's
, remove?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What is the trailing
's
, remove?
StreamReader
is an interface. There're several implementations and the 's
is plural form. I used plural form because the fixes were on several StreamReaders. Let's consult @aweisberg who is native speaker and runs the blog. Ariel, which way is better according to your opinion?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I removed the '
and kept the s
. So now it will look like StreamReader
s
|
||
Hive Changes | ||
____________ | ||
* Fix several memory counting bugs in ``OrcRecordReader`` and ``StreamReader``'s. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What is the trailing 's
, remove?
General Changes | ||
_______________ | ||
* Fix incorrect results from function classification_precision() introduced in release 0.239 | ||
* Add dynamic filtering and bucket pruning support for inner join and semi join, which can avoid full table scan on the probe side of broadcast joins or collocated joins when the build side is small. This can be enabled with the ``experimental.enable-dynamic-filtering`` configuration property and ``enable_dynamic_filtering`` system session property. The build side join key space size can be tuned by the ``experimental.dynamic-filtering-max-per-driver-row-count`` and ``experimental.dynamic-filtering-max-per-driver-size`` configuration properties and ``dynamic_filtering_max_per_driver_row_count`` and ``dynamic_filtering_max_per_driver_size`` session properties. Currently, only Hive connector can benefit from the feature. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
- this item is a performance improve, so the line should starts with "Improve".
Improve performance of broadcast and collocated join by supporting dynamic filtering and bucket pruning for inner and semi joins.
- the item is way too long. Remove every after "The build side join key ...", and add a link to the PR or the issue instead. Something like
(:pr:`12345`)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Improve performance of broadcast and collocated join by supporting dynamic filtering and bucket pruning for inner and semi joins.
@caithagoras thanks for the rewriting. But this PR's primary goal is not improving the join's performance but the scan performance on the probe side. So I changed it to Improve performance for queries with broadcast or collocated joins by adding dynamic filtering and bucket pruning support.
commit message looks good to me, thanks! |
Msg LGTM, thanks! |
5c87163
to
219721c
Compare
Thanks @ajaygeorge @caithagoras for reviewing. I have addressed the comments. Will you be able to take another look? |
219721c
to
94696f0
Compare
LGTM % sphinx compilation error
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
94696f0
to
e184c66
Compare
Missing Release Notes
Daniel Ohayon
George Wang
Vic Zhang
Extracted Release Notes
geometry_nearest_points
to find nearest points of a pair of geometries.hive.enable-parquet-dereference-pushdown
.StageStatistics
andOperatorStatistics
toQueryCompletedEvent
and remove stage and operator statistics fromQueryStatistics
./security/authorization
.experimental.enable-dynamic-filtering
toTrue
to enable the feature. Configsexperimental.dynamic-filtering-max-per-driver-row-count
andexperimental.dynamic-filtering-max-per-driver-size
are available to tune the size on the build side join key space. Currently, only Hive connector can benefit from the feature.ST_Intersection
.All Commits