branch-3.1: [feature](index)Support light index add for inverted index without parser and ngram bf index #52894
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
|
Thank you for your contribution to Apache Doris. Please clearly describe your PR:
|
Member
Author
|
run buildall |
3 similar comments
Member
Author
|
run buildall |
Member
Author
|
run buildall |
Member
Author
|
run buildall |
TPC-H: Total hot run time: 39882 ms |
TPC-DS: Total hot run time: 190249 ms |
ClickBench: Total hot run time: 30.32 s |
Member
Author
|
run buildall |
TPC-H: Total hot run time: 40261 ms |
TPC-DS: Total hot run time: 189846 ms |
ClickBench: Total hot run time: 29.55 s |
Member
Author
|
run buildall |
…pache#48461) Currently, NGram bloom filter index only supports directly schema change, and users need to build indexes incrementally when using it. The design goal is that ngrambf supports light_index_change, including local and cloud mode, which can incrementally add indexes or build indexes on stock data. Inverted indexes are currently only supported in local mode for light_schema_change, cloud mode is still a directly SC, this time it does not involve inverted indexes, and its functionality remains unchanged. After the completion of the function, the NGram BF index construction can be used in the following way, following the existing syntax, does not involve changes or additions. ```sql alter table t1 add index idx_ngram_k2 (`k2`) using ngram_bf properties("bf_size" = "1024", "gram_size" = "3"); create index idx_ngram_k2 (`k2`) on t1 using ngram_bf properties("bf_size" = "1024", "gram_size" = "3"); build index idx_ngram_k2 on t1; show alter table column; cancel build index on t1; ``` **NOTE:** Currently, building an index by partition is not supported. If you want to build an index for stock data, you need to build it for all data, including new data written after the Add index has been added. Build index by partition will be supported in next stage. Support light index change for NGram bf index
…without parser (apache#52251) Problem Summary: This PR adds support for lightweight index add for inverted indexes without parser in cloud mode and extends the existing “light” mode for NGRAM Bloom Filter indexes. And also introduces a new session variable (enable_add_index_for_new_data) to control the above logic.
airborne12
force-pushed
the
pick-index-add-cloud
branch
from
d5d5395 to
f8795dc
Compare
Member
Author
|
run buildall |
TPC-H: Total hot run time: 39521 ms |
TPC-DS: Total hot run time: 196675 ms |
ClickBench: Total hot run time: 30.26 s |
morrySnow
approved these changes
Jul 9, 2025
airborne12
added a commit
to airborne12/apache-doris
that referenced
this pull request
Jul 9, 2025
…x without parser and ngram bf index apache#48461 apache#52251 (apache#52894) cherry pick from apache#48461 and apache#52251 --------- Co-authored-by: qiye <jianliang5669@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
cherry pick from #48461 and #52251