[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Vihang Karajgaonkar has submitted this change and it was merged. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Reviewed-on: http://gerrit.cloudera.org:8080/17298 Tested-by: Impala Public Jenkins Reviewed-by: Vihang Karajgaonkar --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 713 insertions(+), 46 deletions(-) Approvals: Impala Public Jenkins: Verified Vihang Karajgaonkar: Looks good to me, approved -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: merged Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 24 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Sourabh Goyal has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 23: (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/23/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/23/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@3024 PS23, Line 3024: LOG.debug("Couldn't rename " + tableInfo); > Does this need to throw an error? Good catch! Yes we should throw an error here? Will MetaException be fine? I have a follow up patch planned for drop table issue. I can address this comment in that patch -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 23 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 09 Jun 2021 17:38:56 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Vihang Karajgaonkar has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 23: Code-Review+2 (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/23/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/23/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@3024 PS23, Line 3024: LOG.debug("Couldn't rename " + tableInfo); Does this need to throw an error? I know you have had quite a few iterations of this patch, so I am okay doing this as a followup. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 23 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 09 Jun 2021 17:00:47 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Sourabh Goyal has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 23: @Quanlong @Vihang: The latest test run is successful. Please take a look again and share your feedback. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 23 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 09 Jun 2021 01:11:00 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 23: Verified+1 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 23 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 09 Jun 2021 00:25:47 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 22: Verified+1 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 22 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 09 Jun 2021 00:10:13 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 23: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7207/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 23 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 08 Jun 2021 18:33:59 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 22: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7206/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 22 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 08 Jun 2021 18:15:52 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 22: Verified-1 Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7202/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 22 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 08 Jun 2021 01:24:56 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 22: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7202/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 22 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Mon, 07 Jun 2021 19:34:37 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Quanlong Huang has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: > Patch Set 21: > > The failure does not seem to be related to the patch. > > Patch Set 21: Verified-1 > > > > Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7192/ I think we can resove it by a rebase. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Mon, 07 Jun 2021 12:44:11 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7197/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Fri, 04 Jun 2021 03:40:56 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7197/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 21:50:32 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Sourabh Goyal has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: The failure does not seem to be related to the patch. > Patch Set 21: Verified-1 > > Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7192/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 21:48:07 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: Verified-1 Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7192/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 12:37:41 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8840/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 06:56:47 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 21: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7192/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 06:36:50 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#21). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 713 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/21 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 21 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 19: Verified-1 Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7190/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 19 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 01:45:08 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 20: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8837/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 20 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 00:34:56 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 20: (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/20/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/20/tests/custom_cluster/test_metastore_service.py@698 PS20, Line 698: , flake8: E231 missing whitespace after ',' -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 20 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 03 Jun 2021 00:15:17 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#20). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 708 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/20 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 20 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 18: Verified-1 Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7189/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 18 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 23:26:00 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 19: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7190/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 19 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 19:44:07 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 19: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8836/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 19 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 19:42:33 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 19: (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/19/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/19/tests/custom_cluster/test_metastore_service.py@696 PS19, Line 696: , flake8: E231 missing whitespace after ',' -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 19 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 19:22:25 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#19). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 706 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/19 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 19 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 18: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8834/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 18 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 17:51:06 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 18: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7189/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 18 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 17:30:56 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#18). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 605 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/18 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 18 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Sourabh Goyal has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 17: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/17/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/17/tests/custom_cluster/test_metastore_service.py@443 PS17, Line 443: e > flake8: E501 line too long (96 > 90 characters) Ack http://gerrit.cloudera.org:8080/#/c/17298/17/tests/custom_cluster/test_metastore_service.py@444 PS17, Line 444: c > flake8: E501 line too long (93 > 90 characters) Ack -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 17 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 17:28:13 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 17: Verified-1 Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7187/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 17 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 07:13:02 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 17: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7187/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 17 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 01:18:10 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 17: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8828/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 17 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 02 Jun 2021 00:17:56 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 17: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/17/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/17/tests/custom_cluster/test_metastore_service.py@443 PS17, Line 443: e flake8: E501 line too long (96 > 90 characters) http://gerrit.cloudera.org:8080/#/c/17298/17/tests/custom_cluster/test_metastore_service.py@444 PS17, Line 444: c flake8: E501 line too long (93 > 90 characters) -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 17 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 01 Jun 2021 23:57:01 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#17). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 604 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/17 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 17 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 15: Verified-1 Build failed: https://jenkins.impala.io/job/gerrit-verify-dryrun/7183/ -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 15 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 01 Jun 2021 23:39:28 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 16: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8825/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 16 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 01 Jun 2021 19:17:32 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 16: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/16/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/16/tests/custom_cluster/test_metastore_service.py@443 PS16, Line 443: e flake8: E501 line too long (96 > 90 characters) http://gerrit.cloudera.org:8080/#/c/17298/16/tests/custom_cluster/test_metastore_service.py@444 PS16, Line 444: c flake8: E501 line too long (93 > 90 characters) -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 16 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 01 Jun 2021 18:55:42 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#16). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 604 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/16 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 16 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 15: Build started: https://jenkins.impala.io/job/gerrit-verify-dryrun/7183/ DRY_RUN=true -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 15 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 01 Jun 2021 17:46:13 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 15: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8800/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 15 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 26 May 2021 23:42:05 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 15: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/15/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/15/tests/custom_cluster/test_metastore_service.py@443 PS15, Line 443: e flake8: E501 line too long (96 > 90 characters) http://gerrit.cloudera.org:8080/#/c/17298/15/tests/custom_cluster/test_metastore_service.py@444 PS15, Line 444: c flake8: E501 line too long (93 > 90 characters) -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 15 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 26 May 2021 23:21:22 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#15). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 597 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/15 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 15 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 14: Build Failed https://jenkins.impala.io/job/gerrit-code-review-checks/8799/ : Initial code review checks failed. See linked job for details on the failure. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 14 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 26 May 2021 22:35:17 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 14: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/14/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/14/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@2978 PS14, Line 2978: private void renameNonTransactionalTableIfExists(String oldDbNameWithCatalog, String oldTableName, line too long (100 > 90) http://gerrit.cloudera.org:8080/#/c/17298/14/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@2979 PS14, Line 2979:String newDbNameWithCatalog, String newTableName, line too long (100 > 90) -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 14 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Wed, 26 May 2021 22:25:08 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#14). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 596 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/14 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 14 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Vihang Karajgaonkar has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 8: (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/8/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/8/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@843 PS8, Line 843: newTable.getTableName() > Good catch ! Yes, that makes sense to me. I think we should also do it for other variations of the alter table APIs in this class. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 8 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 18:07:28 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Sourabh Goyal has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 13: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/8/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/8/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@843 PS8, Line 843: e > I just realize that ALTER TABLE RENAME TO is also handled here. So I think Good catch ! For alter_table cases, we can instead call renameTable method in CatalogServiceCatalog. The signature of the method is: // Renames a table. Equivalent to an atomic drop + add of the table. public Pair renameTable( TTableName oldTableName, TTableName newTableName) Thoughts? cc - @Vihang http://gerrit.cloudera.org:8080/#/c/17298/8/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/8/tests/custom_cluster/test_metastore_service.py@449 PS8, Line 449: """ > Could you add a test for renaming the table name, and verify that the old t Sure will add a test for rename table. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 13 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 17:27:36 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 13: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8783/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 13 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 01:05:36 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#13). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 512 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/13 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 13 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 12: Build Failed https://jenkins.impala.io/job/gerrit-code-review-checks/8782/ : Initial code review checks failed. See linked job for details on the failure. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 12 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 00:27:03 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 11: Build Failed https://jenkins.impala.io/job/gerrit-code-review-checks/8781/ : Initial code review checks failed. See linked job for details on the failure. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 11 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 00:24:38 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 10: Build Failed https://jenkins.impala.io/job/gerrit-code-review-checks/8780/ : Initial code review checks failed. See linked job for details on the failure. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 10 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 00:22:21 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#12). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 511 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/12 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 12 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#11). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 510 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/11 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 11 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 10: (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/10/be/src/util/backend-gflag-util.cc File be/src/util/backend-gflag-util.cc: http://gerrit.cloudera.org:8080/#/c/17298/10/be/src/util/backend-gflag-util.cc@310 PS10, Line 310: >>> 6b42204e2... IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis line too long (92 > 90) -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 10 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 00:12:59 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#10). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 519 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/10 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 10 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 9: Build Failed https://jenkins.impala.io/job/gerrit-code-review-checks/8779/ : Initial code review checks failed. See linked job for details on the failure. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 9 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Tue, 25 May 2021 00:04:33 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 9: (1 comment) http://gerrit.cloudera.org:8080/#/c/17298/9/be/src/util/backend-gflag-util.cc File be/src/util/backend-gflag-util.cc: http://gerrit.cloudera.org:8080/#/c/17298/9/be/src/util/backend-gflag-util.cc@310 PS9, Line 310: >>> 6b42204e2... IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis line too long (92 > 90) -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 9 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Mon, 24 May 2021 23:57:51 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#9). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 519 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/9 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 9 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Quanlong Huang has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 8: (2 comments) http://gerrit.cloudera.org:8080/#/c/17298/8/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/8/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@843 PS8, Line 843: newTable.getTableName() I just realize that ALTER TABLE RENAME TO is also handled here. So I think we should also call this on the original table name if it differs from the new table. Otherwise, the old table will still exist. http://gerrit.cloudera.org:8080/#/c/17298/8/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/8/tests/custom_cluster/test_metastore_service.py@449 PS8, Line 449: """ Could you add a test for renaming the table name, and verify that the old table disappear and the new table appear? -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 8 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Thu, 29 Apr 2021 07:49:28 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Impala Public Jenkins has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 8: Build Successful https://jenkins.impala.io/job/gerrit-code-review-checks/8643/ : Initial code review checks passed. Use gerrit-verify-dryrun-external or gerrit-verify-dryrun to run full precommit tests. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 8 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Mon, 26 Apr 2021 20:03:43 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Vihang Karajgaonkar has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 8: Thanks. The changes looks good to me. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 8 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Mon, 26 Apr 2021 20:02:36 + Gerrit-HasComments: No
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Sourabh Goyal has posted comments on this change. ( http://gerrit.cloudera.org:8080/17298 ) Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. Patch Set 8: (4 comments) http://gerrit.cloudera.org:8080/#/c/17298/5//COMMIT_MSG Commit Message: http://gerrit.cloudera.org:8080/#/c/17298/5//COMMIT_MSG@9 PS5, Line 9: For transactional tables, catalogd already guarantees consitent table : metadata reads > nit, Can you please reformat this commit msg to 72 line width as per the co Ack http://gerrit.cloudera.org:8080/#/c/17298/7/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java File fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java: http://gerrit.cloudera.org:8080/#/c/17298/7/fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java@2966 PS7, Line 2966: n > I found that removing the table directly from catalog_ doesn't take the met Sure. http://gerrit.cloudera.org:8080/#/c/17298/7/tests/custom_cluster/test_metastore_service.py File tests/custom_cluster/test_metastore_service.py: http://gerrit.cloudera.org:8080/#/c/17298/7/tests/custom_cluster/test_metastore_service.py@439 PS7, Line 439: invalid > nit, s/removed/invalidated Ack http://gerrit.cloudera.org:8080/#/c/17298/7/tests/custom_cluster/test_metastore_service.py@442 PS7, Line 442: removed > nit, s/removed/invalidated For drop case, we remove (and not invalidate) from the cache. -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 8 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar Gerrit-Comment-Date: Mon, 26 Apr 2021 19:43:26 + Gerrit-HasComments: Yes
[Impala-ASF-CR] IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis
Hello Quanlong Huang, Vihang Karajgaonkar, Impala Public Jenkins, I'd like you to reexamine a change. Please visit http://gerrit.cloudera.org:8080/17298 to look at the new patch set (#8). Change subject: IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis .. IMPALA-10648: Invalidate catalogd table metadata cache for HMS DDL apis For transactional tables, catalogd already guarantees consitent table metadata reads based on the writeIdList passed in the request. For non transactional tables, the reads are eventually consistent as in event processor thread in the background, processes HMS events for the table and updates its metadata. In this patch, to ensure strong consistency guarantees for external tables,we invalidate the table metadata from cache if HMS DDL apis like alter/drop table/partition are accessed from catalogd's metastore server. As a result of which, any subsequent get table request fetches the table from HMS and loads it in cache. This ensures that any get_table/get_partition requests after DDL operations on same table return updated table metadata. This behavior has a performance penalty since metadata loading in cache takes time specially for large tables. The change is behind catalogd server's flag: invalidate_hms_cache_on_ddls which is enabled by default. The flag needs to be turned off in case of a performance bottleneck. Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 --- M be/src/catalog/catalog-server.cc M be/src/util/backend-gflag-util.cc M common/thrift/BackendGflags.thrift M fe/src/main/java/org/apache/impala/catalog/metastore/MetastoreServiceHandler.java M fe/src/main/java/org/apache/impala/service/BackendConfig.java M tests/custom_cluster/test_metastore_service.py 6 files changed, 517 insertions(+), 46 deletions(-) git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/98/17298/8 -- To view, visit http://gerrit.cloudera.org:8080/17298 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: newpatchset Gerrit-Change-Id: Idb9cc22ebfb51948433e4d57f4705ce201acaf98 Gerrit-Change-Number: 17298 Gerrit-PatchSet: 8 Gerrit-Owner: Sourabh Goyal Gerrit-Reviewer: Impala Public Jenkins Gerrit-Reviewer: Quanlong Huang Gerrit-Reviewer: Sourabh Goyal Gerrit-Reviewer: Vihang Karajgaonkar