-
Notifications
You must be signed in to change notification settings - Fork 298
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
[HotFix] Add table filter for catalog properties #2310
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #2310 +/- ##
============================================
+ Coverage 53.39% 53.53% +0.13%
- Complexity 4354 4369 +15
============================================
Files 517 517
Lines 29856 29872 +16
Branches 2908 2910 +2
============================================
+ Hits 15942 15991 +49
+ Misses 12637 12594 -43
- Partials 1277 1287 +10
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
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.
BasicMixedIcebergCatalog also need this.
I think table-filter should be a generic feature for the external catalog. Maybe you can help confirm if it is feasible. @HuangFru |
What do you mean? Judging from the current code structure, it should be effective for all external catalogs. |
Why? For external Mix-iceberg? |
Yes, but I find a |
Table filter is only for external catalog, Unified includes internal and external |
I got you. |
ams/server/src/main/java/com/netease/arctic/server/catalog/ExternalCatalog.java
Outdated
Show resolved
Hide resolved
Table filter works for AMS, but InternalCatalog will not use UnifiedCatalog for ServerCatalog implementation. |
The catalog type[Internal/External] is a concept of AMS Instance, for engine side, all catalogs have no different. |
So you mean the engine side also needs the table filter? |
…-table-filter # Conflicts: # core/src/main/java/com/netease/arctic/mixed/BasicMixedIcebergCatalog.java
AFAIK, the table filter feature helps AMS filter out tables that are not relevant for attention, and currently, only the external catalogs require it. Although the HDYT? @HuangFru @wangtaohz @baiyangtx |
There is indeed no scenario on the engine side that requires filtering tables or databases, I'll rollback the change. |
ams/server/src/test/java/com/netease/arctic/server/catalog/TestServerCatalog.java
Outdated
Show resolved
Hide resolved
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.
Should table-filter
needs to be applied to the listDatabase
interface ?
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
I think it's hard to cut a db filter from the table filter to apply to the |
Have you considered restoring the previous parameter |
I don’t know if the community felt that table filter could replace |
IMO, it seems unnecessary to filter databases now, and the current implementation of this PR is sufficient. If there is a need to filter databases, it can be addressed in another issue. |
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, This parameter was originally introduced to reduce the number of tables processed by AMS to improve stability. From this perspective, listDatabase is unnecessary.
* re add table filter * implement in external catalog * add ut case * fix comment * fix comment * fix comment * fix ut * fix update properties * roll back the engine side's filter * resolve conflicts * add ut --------- Co-authored-by: baiyangtx <xiangnebula@163.com> Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com>
* [AMORO-1812] support spark-based external optimizer * resolve code style error * [AMORO-1951] Support parallelized planning in one optimizer group (#2282) * [AMORO-1951] Support parallelized planning in one optimizer group * [AMORO-1951] add unit test for OptimizingQueue and DefaultOptimizingService * [AMORO-1951] optimize default parameters * fix bugs * fix warnings and spotless issues * merge from #2290 * add apache license and fix spotless * fix config error * Update ams/server/src/main/java/com/netease/arctic/server/DefaultOptimizingService.java Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> * add annotations * fix compile errors * fix import problem * remove isDebugEnabled() * spotless apply * Update ArcticManagementConf.java * fix reboot bug and supply document content * use MoreObjects.toStringHelper for OptimizerThread.java * Merged from [AMORO-2376] Print right log info after calculating and sorting tables * fix import problem * remove unused codes * spotless * remove incorrect comments * add max-planning-parallelism to config --------- Co-authored-by: majin1102 <majin1102@163.com> Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> * [AMORO-2378] The optimizer based on Flink on YARN should prioritize loading the optimizer-job.jar (#2379) * load optimizer jar first * fix code style * change config name * add config taskmanager.memory.managed.fraction * fix * [AMORO-2222] [Improvement]: Skip cleaning up dangling delete files for Iceberg V1 table (#2361) * [AMORO-2222] [Improvement]: Skip cleaning up dangling delete files for Iceberg V1 table * Update IcebergTableMaintainer.java The `total-delete-files` could be 0. --------- Co-authored-by: wangtaohz <103108928+wangtaohz@users.noreply.github.com> * [AMORO-2404] fix Mixed Hive table mistakenly deletes hive files during expiring snapshots (#2405) get hive locations return the uri path * [AMORO-2407] Fix access data file from dashboard of non-partitioned table (#2408) * fix null partition * fix listing files of non-partitioned iceberg table * [AMORO-2383] Add serialVersionUID to RewriteFilesInput (#2384) * add serialVersionUID * fix comment * [AMORO-1720] Fix Mixed Format KeyedTable expiring all the snapshots with optimized sequence (#2394) * should not expire the latest snapshot contains optimized sequence * add visible for testing * add fetchLatestNonOptimizedSnapshotTime for base store * get hive locations return the uri path * refactor codes and fix comments * improve for exclude files is empty for expring snapshots --------- Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> * [AMORO-2386][AMS] Configure `iceberg.worker.num-threads` in the config.yaml (#2393) * [AMORO-2386][AMS] Configure `iceberg.worker.num-threads` in the config.yaml * Fix * [AMORO-2386][AMS] reuse config `table-manifest-io.thread-count` and reuse thread pool * Add comment * [AMORO-1716] [Improvement]: sort the table list returned by server (#2362) * improve: sort the table list returned by server * optimize: sort tables by format * optimize: optimiz tables sorting * style: udpate comment --------- Co-authored-by: chenyuzhi <chenyuzhi@corp.netease.com> Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> * [HotFix] Re-add table-filter to Server ExternalCatalog (#2310) * re add table filter * implement in external catalog * add ut case * fix comment * fix comment * fix comment * fix ut * fix update properties * roll back the engine side's filter * resolve conflicts * add ut --------- Co-authored-by: baiyangtx <xiangnebula@163.com> Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> * [AMORO-2299]: Cancel the running optimizing process from ams web console (#2297) * cancel the running opimizing process from ams web console * refact code to avoid NPE * add o comment for com.netease.arctic.server.table.TableService#getServerTableIdentifier * change the cancel post api to be more restful style * [AMORO-2415] Print GC date stamps (#2416) add gc timestamp * Update wrong comments in SnapshotsExpiringExecutor.java (#2422) * [AMORO-2276]: UnifiiedCatalog for Spark Engine (#2269) * Add UnifiedSparkCatalog under spark common module * Extract MixedSparkCatalogBase and MixedSparkSessionCatalogBase to spark common module * Refactor spark unit test framework to adapt unifed catalog tests and mixed format tests. * [AMORO-2261] Extract the deleting dangling files from the cleaning orphan files (#2403) * [Improvement]: Extract the deleting dangling files from the cleaning orphan files * [Improvement]: Extract the deleting dangling files from the cleaning orphan files * [Improvement]: Extract the deleting dangling files from the cleaning orphan files * [AMORO-1341] [Flink]: Support UnifiedCatalog to contain Mixed format table in Flink Engine (#2370) * [AMORO-1341] [Flink]: Support UnifiedCatalog to contain Mixed format table in Flink Engine * [AMORO-2413] Need to select the first db after switching to another Catalog (#2419) * fix: If the current catalog is not the one in the query, the first db is selected by default. * build dashboard frontend --------- Co-authored-by: wangtao <wangtao3@corp.netease.com> * [HotFix] Fix loading the optimizing snapshot id of change store for Mixed Format KeyedTable (#2430) fix load target change snapshot id * [AMORO-2260] Show the format version of iceberg table (#2425) [AMORO-2260] Show the format version of Iceberg Table Signed-off-by: tcodehuber <tcodehuber@gmail.com> * [AMORO-2115] Support displaying Optimizing tasks (#2322) * dashboard: rename optimized to optimizing * dashboard: support optimizing taskes * add optimizer token * dashboard: modify column width * dashboard: build * sort the metrics field and change record cnt to long * modify MetricsSummary Compatibility * dashbard: build * Update ams/server/src/main/java/com/netease/arctic/server/optimizing/TaskRuntime.java Co-authored-by: Qishang Zhong <zhongqishang@gmail.com> * fix * support input metrics and output metrics for optimizing process * dashboard: support optimizing metrics * dashbard: build * dashboard:rebuild * support MetricsSummary to map * optimizing task supports input output * dashboard: optimizing tasks support input and output * dashboard: not display seconds when longer than 1 hour * dashboard: optimizing process show summary * remove useless import * dashboard: build * as head * dashbard: build * change process status to CLOSED after cancel process * remove useless log * dashboard: refresh after cancelled * support cancel optimizing tasks * dashboard: handle exception when can't cancel optimizing process * throw exception when can't cancel optimizing process * dashboard: build * dashboard: refresh optimizing process when exist optimizing detail page * dashboard: build * fix cost time is 0ms * change metrics name * fix task startTime and endTime * fix costTime * using Preconditions.checkArgument * fix task reset * add comments * cancel tasks before closing optimizing process * fix unit test * fix cancel task * as head * Revert "as head" This reverts commit e469e71. * dashboard: build --------- Co-authored-by: Qishang Zhong <zhongqishang@gmail.com> * [AMORO-2385] Make the maximum input file size for per optimize thread configurable (#2387) * add config self-optimizing.max-input-file-size-per-thread * add doc * add resource group property max-input-file-size-per-thread * add doc * fix compile * [Hotfix] Add database filter to Server ExternalCatalog (#2414) * [Hotfix] Add database filter to Server ExternalCatalog * [Hotfix] Add database filter to Server ExternalCatalog * Rename config database.filter-regular-expression to database-filter --------- Co-authored-by: baiyangtx <xiangnebula@163.com> * [AMORO-2423] [Flink]: Using 'mixed_iceberg' and 'mixed_hive' indentifier to CREATE CATALOG and deprecate 'arctic' identifier (#2424) * [AMORO-2423] [Flink]: Using 'mixed_iceberg' and 'mixed_hive' identifiers to CREATE CATALOG and deprecate 'arctic' identifier * [AMORO-2316] The Files page supports filtering by partition name and sorting by dictionary value. (#2420) * AMORO-2316: The Files page supports filtering by partition name and sorting by dictionary value. * build dashboard frontend * dashboard: build --------- Co-authored-by: wangtao <wangtao3@corp.netease.com> * [AMORO-1892] Improve the SQL Shortcuts in Terminal Web UI (#2434) * [AMORO-1892] Improve the SQL Shortcuts in Terminal Web UI * refactor some code --------- Signed-off-by: tcodehuber <tcodehuber@gmail.com> * [AMORO-2440] Fix the batch deletion of Change Store files for Mixed Format Table (#2439) fix remove change files * [AMORO-2418] Exclude kryo dependency from flink-optimizer (#2437) exclude kryo * [AMORO-2441] Fix `TableEntriesScan` when no file format is specified in the file name (#2442) * fix TableEntriesScan without format suffix * using the file format from entries * [Hotfix] Fix fetchLatestNonOptimizedSnapshotTime (#2396) * fix fetchLatestNonOptimizedSnapshotTime * fix * spotless * fix ut * rename data-expire.since * using UTC zone for snapshot timestamp * resolve conflict * rerview * spotless * review * review * [AMORO-2344] [Flink]: Support UnifiedCatalog to contain Iceberg format table in Flink Engine (#2427) * [AMORO-2330] Improve major plan (#2332) * [AMORO-2330][AMS] Improve major plan * [AMORO-2330][AMS] remove filter dataFileWith1Pos * Fix comments * Fix * Move the rollback logic of undersized segment to the split task stage * Rename * Fix * Rollback method name `fileShouldRewrite` * Rollback mixed table format full trigger condition & reuse `isUndersizedSegmentFile` * Rollback testSegmentFilesBase() * TreeNodeTaskSplitter logic keep same with bin-pack * Improve code duplicate --------- Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> * [AMORO-1810] Check the validity of the heatbeat interval when an opti… (#2432) * [AMORO-1810] Check the validity of the heatbeat interval when an optimizer start * adjust the import way * resolve some logic code * refactor code * refactor code * fix ut error * resolve ut error * fix ut error * fix ut error * fix ut error * fix ci error * refactor code * refactor code * refactor code * [AMORO-1812] support spark-based external optimizer * resolve code style error * refactor code * refactor code * bugfix * refactor code * refactor code * code style * bugfix * bugfix --------- Signed-off-by: tcodehuber <tcodehuber@gmail.com> Co-authored-by: JinMat <majin1102@gmail.com> Co-authored-by: majin1102 <majin1102@163.com> Co-authored-by: ZhouJinsong <zhoujinsong0505@163.com> Co-authored-by: wangzeyu <hameizi369@gmail.com> Co-authored-by: ConradJam <jam.gzczy@gmail.com> Co-authored-by: wangtaohz <103108928+wangtaohz@users.noreply.github.com> Co-authored-by: yeatsliao <liaoyt66066@gmail.com> Co-authored-by: Qishang Zhong <zhongqishang@gmail.com> Co-authored-by: chenyuzhi459 <553673833@qq.com> Co-authored-by: chenyuzhi <chenyuzhi@corp.netease.com> Co-authored-by: HuangFru <68625618+HuangFru@users.noreply.github.com> Co-authored-by: baiyangtx <xiangnebula@163.com> Co-authored-by: xujiangfeng001 <104614523+xujiangfeng001@users.noreply.github.com> Co-authored-by: Xianxun Ye <yesorno828423@gmail.com> Co-authored-by: liuweimin <minteliu.l@gmail.com> Co-authored-by: wangtao <wangtao3@corp.netease.com> Co-authored-by: Xavier Bai <xuba@cisco.com>
Why are the changes needed?
We've lost table-filter in #2232.
Brief change log
How was this patch tested?
Add some test cases that check the changes thoroughly including negative and positive cases if possible
Add screenshots for manual tests if appropriate
Run test locally before making a pull request
Documentation