Related to #31293
Implement QueryIterator for the Go SDK to enable efficient iteration
over large query result sets using PK-based pagination.
Key changes:
- Add QueryIterator interface and implementation with PK-based
pagination
- Support Int64 and VarChar primary key types for pagination filtering
- Add QueryIteratorOption with batchSize, limit, filter, outputFields
config
- Fix ResultSet.Slice to handle Query results without IDs/Scores
- Add comprehensive unit tests and integration tests
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: the iterator requires the collection primary key (PK)
to be present in outputFields so PK-based pagination and accurate row
counting work. The constructor enforces this by appending the PK to
outputFields when absent, and all pagination (lastPK tracking, PK-range
filters) and ResultCount calculations depend on that guaranteed PK
column.
- New capability: adds a public QueryIterator API (Client.QueryIterator,
QueryIterator interface, QueryIteratorOption) that issues server-side
Query RPCs in configurable batches and implements PK-based pagination
supporting Int64 and VarChar PKs, with options for batchSize, limit,
filter, outputFields and an upfront first-batch validation to fail fast
on invalid params.
- Removed/simplified logic: ResultSet.Slice no longer assumes IDs and
Scores are always present — it branches on presence of IDs (use IDs
length when non-nil; otherwise derive row count from Fields[0]) and
guards Scores slicing. This eliminates redundant/unsafe assumptions and
centralizes correct row-count logic based on actual returned fields.
- No data loss or behavior regression: pagination composes the user
filter with a PK-range filter and always requests the PK field, so
lastPK is extracted from a real column and fetchNextBatch only advances
when rows are returned; EOF is returned only when the server returns no
rows or iterator limit is reached. ResultSet.Slice guards prevent panics
for queries that lack IDs/Scores; Query RPC → ResultSet.Fields remains
the authoritative path for row data, so rows are not dropped and
existing query behavior is preserved.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
Support crate analyzer with file resource info, and return used file
resource ids when validate analyzer.
Save the related resource ids in collection schema.
relate: https://github.com/milvus-io/milvus/issues/43687
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: analyzer file-resource resolution is deterministic and
traceable by threading a FileResourcePathHelper (collecting used
resource IDs in a HashSet) through all tokenizer/analyzer construction
and validation paths; validate_analyzer(params, extra_info) returns the
collected Vec<i64) which is propagated through C/Rust/Go layers to
callers (CValidateResult → RustResult::from_vec_i64 → Go []int64 →
querypb.ValidateAnalyzerResponse.ResourceIds →
CollectionSchema.FileResourceIds).
- Logic removed/simplified: ad‑hoc, scattered resource-path lookups and
per-filter file helpers (e.g., read_synonyms_file and other inline
file-reading logic) were consolidated into ResourceInfo +
FileResourcePathHelper and a centralized get_resource_path(helper, ...)
API; filter/tokenizer builder APIs now accept &mut
FileResourcePathHelper so all file path resolution and ID collection use
the same path and bookkeeping logic (redundant duplicated lookups
removed).
- Why no data loss or behavior regression: changes are additive and
default-preserving — existing call sites pass extra_info = "" so
analyzer creation/validation behavior and error paths remain unchanged;
new Collection.FileResourceIds is populated from resp.ResourceIds in
validateSchema and round‑tripped through marshal/unmarshal
(model.Collection ↔ schemapb.CollectionSchema) so schema persistence
uses the new list without overwriting other schema fields; proto change
adds a repeated field (resource_ids) which is wire‑compatible (older
clients ignore extra field). Concrete code paths: analyzer creation
still uses create_analyzer (now with extra_info ""), tokenizer
validation still returns errors as before but now also returns IDs via
CValidateResult/RustResult, and rootcoord.validateSchema assigns
resp.ResourceIds → schema.FileResourceIds.
- New capability added: end‑to‑end discovery, return, and persistence of
file resource IDs used by analyzers — validate flows now return resource
IDs and the system stores them in collection schema (affects tantivy
analyzer binding, canalyzer C bindings, internal/util analyzer APIs,
querynode ValidateAnalyzer response, and rootcoord/create_collection
flow).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
issue: https://github.com/milvus-io/milvus/issues/45919
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: FlushAll verification must accept both per-channel
FlushAllTss (new schema) and the legacy single FlushAllTs;
GetFlushAllState chooses the verification path based on which field is
present and treats a channel as flushed only if its channel checkpoint
timestamp >= the applicable threshold (per-channel timestamp or legacy
FlushAllTs).
- Logic removed/simplified: The previous mixed/ambiguous checks were
split into two focused
routines—verifyFlushAllStateByChannelFlushAllTs(logger, channel,
flushAllTss) and verifyFlushAllStateByLegacyFlushAllTs(logger, channel,
flushAllTs)—and GetFlushAllState now selects one path. This centralizes
compatibility logic, eliminates interleaved/duplicated checks, and
retains the outer-loop short-circuiting on the first unflushed channel.
- Why this does NOT cause data loss or regressions: Changes only affect
read-only verification paths (GetFlushAllState/GetFlushState) that
compare in-memory channel checkpoints (meta.GetChannelCheckpoint) to
provided thresholds; no writes to checkpoints or persisted state occur
and FlushAll enqueue/wait behavior is unchanged. Unit tests were added
to cover legacy FlushAllTs behavior and the new FlushAllMsgs→FlushAllTs
extraction, exercising both code paths.
- Enhancement scope and location: Adds backward-compatible support and
concrete FlushAllTs extraction from streaming FlushAllMsgs in Proxy
(internal/proxy/task_flush_all_streaming.go) and compatibility verifiers
in DataCoord (internal/datacoord/services.go), plus corresponding tests
(internal/datacoord/services_test.go, internal/proxy/*_test.go).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
issue: #46576
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: During meta load, only tasks that are truly
terminal-cleaned (states cleaned or unknown) should be dropped; all
other non-terminal tasks (including timeout and completed) must be
restored so the inspector can reattach them to executing/cleaning queues
and finish their cleanup lifecycle.
- Removed/simplified logic: loadMeta no longer uses the broad
isCompactionTaskFinished predicate (which treated timeout, completed,
cleaned, unknown as terminal). It now uses the new
isCompactionTaskCleaned predicate that only treats cleaned/unknown as
terminal. This removes the redundant exclusion of timeout/completed
tasks and simplifies the guard to drop only cleaned/unknown tasks.
- Bug fix (root cause & exact change): Fixes issue #46576 — the previous
isCompactionTaskFinished caused timeout/completed tasks to be skipped
during meta load and thus not passed into restoreTask(). The PR adds
isCompactionTaskCleaned and replaces the finished check so timeout and
completed tasks are included in restoreTask() and re-attached to the
inspector’s existing executing/cleaning queues.
- No data loss or regression: Tasks in cleaned/unknown remain dropped
(isCompactionTaskCleaned still returns true for cleaned/unknown).
Non-terminal timeout/completed tasks now follow the same restoreTask()
control path used previously for restored tasks — they are enqueued into
the inspector’s queue/executing/cleaning flows rather than being
discarded. No exported signatures changed and all restored tasks flow
into existing handlers, avoiding behavior regression or data loss.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
issue: https://github.com/milvus-io/milvus/issues/46116
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: the metric CDCLastReplicatedTimeTick must reflect the
most recent time-tick when replication has effectively processed all
pending messages (including idle periods), so reported replicate lag =
confirmed WAL tick − last replicated tick can reach zero when the server
is idle.
- Exact fix (bug): addresses issue #46116 by ensuring the
last-replicated metric is updated when the server is idle. Concretely, a
new ReplicateMetrics.UpdateLastReplicatedTimeTick(ts uint64) was added
and called from OnConfirmed (OnConfirmed now delegates to
UpdateLastReplicatedTimeTick(msg.TimeTick())), and from Replicate’s
self-controlled-message path when the pending queue is empty — so the
code records the time tick before returning ErrReplicateIgnored.
- Logic simplified / removed: direct, ad-hoc metric writes in
OnConfirmed were replaced by a single UpdateLastReplicatedTimeTick
helper on the metrics implementation. The scattered manual set of
CDCLastReplicatedTimeTick is consolidated into one method, removing
redundant direct metric manipulations and centralizing timestamp
conversion (tsoutil.PhysicalTimeSeconds).
- No data loss / no behavior regression: this change only updates
monitoring metrics and does not alter replication control flow or
message processing. Replicate still returns ErrReplicateIgnored for
self-controlled messages and does not change message persistence or
acknowledgement paths; OnConfirmed continues to be invoked on confirmed
messages but now delegates metric recording to the new method. Therefore
no replication state, message ordering, or persistence semantics are
modified.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
### **User description**
issue: #46367
___
### **PR Type**
Bug fix, Tests
___
### **Description**
- Fix unstable test case by adjusting float precision
- Change listMix float value from 1.1 to 1.111
- Improves test stability for json_contains_any query
___
### Diagram Walkthrough
```mermaid
flowchart LR
A["Test Data Generation"] -- "Adjust float precision" --> B["listMix field value"]
B -- "1.1 to 1.111" --> C["Improved test stability"]
```
<details><summary><h3>File Walkthrough</h3></summary>
<table><thead><tr><th></th><th align="left">Relevant
files</th></tr></thead><tbody><tr><td><strong>Bug
fix</strong></td><td><table>
<tr>
<td>
<details>
<summary><strong>test_milvus_client_query.py</strong><dd><code>Adjust
float precision in test data</code>
</dd></summary>
<hr>
tests/python_client/milvus_client/test_milvus_client_query.py
<ul><li>Modified test data generation in
<br><code>test_milvus_client_query_expr_all_datatype_json_contains_all</code>
method<br> <li> Changed <code>listMix</code> field float value from
<code>1.1</code> to <code>1.111</code> for improved <br>precision<br>
<li> Addresses test instability issue by adjusting floating-point test
data</ul>
</details>
</td>
<td><a
href="https://github.com/milvus-io/milvus/pull/46506/files#diff-d6fe357e4678415bc62596b802571043fa571c7d1b8e841aa43124437dd2f739">+1/-1</a>
</td>
</tr>
</table></td></tr></tbody></table>
</details>
___
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: the test assumes stable float equality/containment
behavior for JSON-typed fields when generating test rows; small changes
in stored float precision can make json_contains_any assertions flaky.
- Exact fix for the bug (refs #46367): in
tests/python_client/milvus_client/test_milvus_client_query.py, the test
data value for the second element of the "listMix" JSON field was
adjusted from i * 1.1 to i * 1.111 in
test_milvus_client_query_expr_all_datatype_json_contains_all to increase
numeric precision and remove instability in json_contains_any
assertions.
- Logic removed/simplified: no production logic was changed or removed —
only a one-line test-data change. There is no control-flow or
algorithmic simplification because the test’s intent and checks remain
identical; the change removes the redundant dependence on a borderline
float value that caused flakiness.
- No data loss or behavior regression: this change only updates
test-generated input
(test_milvus_client_query_expr_all_datatype_json_contains_all) and does
not touch any library or runtime code paths. Production code paths
(query parsing/execution, JSON handling) are unchanged, so no persisted
data, API behavior, or client logic is affected.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: nico <cheng.yuan@zilliz.com>
issue: #46352
- Add GenerateChannelBalancedPrimaryKeys function supporting Int64 and
VarChar PK types with even distribution across channels
- Add GenerateBalancedInt64PKs and GenerateBalancedVarCharPKs helper
functions using murmur3 and crc32 hash algorithms respectively
- Add PrimaryKeyConfig struct to configure PK generation in
InsertAndFlush
- Update InsertAndFlush to use channel balanced PKs instead of random
hash keys for better test coverage
- Add comprehensive unit tests including end-to-end validation with
typeutil.HashPK2Channels to verify exact channel distribution
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Enhanced integration tests with configurable primary-key sequencing
across insert batches.
* Added utilities to generate channel-balanced primary keys for integer
and string types.
* Expanded test coverage validating balanced distribution, uniqueness,
continuation, and large-scale behavior across multiple channels.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: wayblink <anyang.wang@zilliz.com>
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
Related to #46600
The test previously checked if all 3 collection IDs were batched
together in a single InvalidateShardLeaderCache call. This caused
flakiness because the observer may split events across multiple calls.
Fix by accumulating all collection IDs across multiple calls and
verifying that eventually all expected IDs (1, 2, 3) are processed.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: the test asserts that all registered collection IDs
{1,2,3} are eventually processed by InvalidateShardLeaderCache across
any number of calls — i.e., the observer must invalidate every
registered collection ID, not necessarily in a single batched RPC (fixes
flaky assumption from issue #46600).
- Logic removed/simplified: the strict expectation that all three IDs
arrive in one InvalidateShardLeaderCache call was replaced by
accumulating IDs into a ConcurrentSet (collectionIDs.Upsert in the mock)
and asserting eventual containment of 1,2,3. This removes the brittle
per-call batching assertion and uses a set-based accumulation (lines
where the mock calls Upsert and final Eventually checks
collectionIDs.Contain(...)).
- Why this is safe (no data loss or behavior regression): only test
assertions changed — production code (LeaderCacheObserver calling
InvalidateShardLeaderCache) is unchanged. The mock intercepts
InvalidateShardLeaderCache and accumulates req.GetCollectionIDs(); the
test still verifies single-ID handling via the existing len==1 &&
lo.Contains(... ) check (first mock block) and verifies that all IDs
were invalidated over time in the batch scenario (second mock block). No
production code paths were modified, so invalidation behavior and RPC
usage remain identical.
- Bug-fix note: this is a targeted test-only fix for issue #46600 — it
tolerates legitimate splitting of events across multiple
InvalidateShardLeaderCache invocations by aggregating IDs across calls
in the test mock, eliminating flakiness without altering runtime
behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
When computing load diff, binlogs in v1/legacy format have empty
child_fields. In this case, the field_id itself should be used as the
child_id (group_id == field_id for legacy format).
Without this fix, legacy format binlogs are not recognized during diff
computation, causing segments to fail loading and TestProxy to timeout.
Changes:
- Add fallback to use fieldid as child_id when child_fields is empty
- Add LoadDiff::ToString() for debugging
- Add logging for diff in Load/Reopen operations
- Add comprehensive unit tests for legacy format handling
Related to #46594
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: load-diff computation must enumerate every binlog
child group for a field so current vs new segment state comparisons
include all column-group/binlog groups; for legacy (v1) binlogs that
have empty child_fields, the code must treat group_id == field_id to
preserve that mapping.
- Bug fix (resolves#46594): SegmentLoadInfo now normalizes
field_binlog.child_fields() into a vector and falls back to using
field_id as the single child group when child_fields is empty; the same
normalization is applied for both current and new-info paths, ensuring
legacy v1 binlogs are discovered and included in Load/ComputeDiff
results so segments load correctly.
- Logic simplified: removed the implicit assumption that child_fields is
always present by centralizing a single normalization/fallback step used
symmetrically for both diff paths, avoiding ad-hoc special-casing and
unifying iteration over child groups.
- No data loss / no behavior regression: the fallback only activates
when child_fields is empty — non-legacy binlogs continue to use their
child_fields unchanged. Add/drop semantics are preserved because the
same normalization is applied to both sides of the diff. Unit tests
(v1-only, v4-only, mixed cases) were added to validate correctness;
LoadDiff::ToString() and extra logging are diagnostic only.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Co-authored-by: Cai Zhang <cai.zhang@zilliz.com>
---------
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
Related to #46594
Remove the temporary config override that forced EnableStorageV2 to
false in TestProxy. This override caused test failures with the new load
logic, as segments could not be loaded with v1 format.
This PR is a quick fix to make ut back to normal
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
### **Description**
- Add `-buildvcs=false` flag to Go test commands in coverage script
- Increase default session TTL from 10s to 15s
- Update SessionTTL parameter default value from 30 to 15
Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
Co-authored-by: bigsheeper <yihao.dai@zilliz.com>
Co-authored-by: chyezh <chyezh@outlook.com>
Co-authored-by: czs007 <zhenshan.cao@zilliz.com>
issue: https://github.com/milvus-io/milvus/issues/44399
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Release Notes
* **New Features**
* Enhanced pattern matching for string indexes with support for prefix,
postfix, inner, and regex-based matching operations.
* Optimized pattern matching performance through prefix-based filtering
and range-based lookups.
* **Tests**
* Added comprehensive test coverage for pattern matching functionality
across multiple index implementations.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
relate: https://github.com/milvus-io/milvus/issues/46571
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
- Core invariant: the LexicalHighlighter API now expects the match
queries under the parameter name highlight_query (not queries); all call
sites must pass highlight_query to supply match data. This PR assumes
the underlying highlighter behavior and processing of those query values
are unchanged.
- Logic simplified/removed: removed the legacy keyword queries in tests
and updated calls to use highlight_query
(tests/python_client/milvus_client/test_milvus_client_highlighter.py).
This eliminates a redundant/incorrect keyword alias and aligns tests
with the consolidated LexicalHighlighter constructor parameter name.
- Why this does NOT introduce data loss or behavior regression: the
change is a parameter-name rename only — no parsing, matching, or
storage logic was modified. Tests now construct LexicalHighlighter with
pre_tags/post_tags/highlight_search_text/fragment_* and pass the query
list under highlight_query; the highlighter execution path
(client.search → highlighter processing → result['highlight']) is
untouched, so existing highlight outputs and stored data remain
unchanged.
- Other changes: bumped pymilvus test dependency to 2.7.0rc93 in
tests/python_client/requirements.txt to match the updated constructor
signature; scope of change is limited to tests and dependency pinning
(no production code changes).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
Related to #46133
Move jemalloc_stats.go and its test file from pkg/util/hardware to
internal/util/segcore. This is a more appropriate location because:
- jemalloc_stats depends on milvus_core C++ library via cgo
- The pkg directory should remain independent of internal C++
dependencies
- segcore is the natural home for core memory allocator utilities
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Refactor**
* Improved internal code organization by reorganizing memory statistics
collection infrastructure for better maintainability and modularity. No
impact on end-user functionality or behavior.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
relate: https://github.com/milvus-io/milvus/issues/43687
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* File resource operations (add, remove, list) are now unavailable and
return a not-implemented response.
* **Tests**
* Tests updated to expect error responses for those file resource
operations and removed some previous coordination-path assertions.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
issue: #46400
- add new server label rule.
- add `MILVUS_SERVER_LABEL_RESOURCE_GROUP` to determine the resource
group of querynode.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Automatic creation of resource groups when nodes with resource-group
labels join.
* Expanded server-label system supporting role-specific and global
labels.
* **Bug Fixes**
* Node acceptance now enforces resource-group name compatibility,
preventing cross-group assignment.
* **Refactor**
* Node handling flows updated to use richer node information for
assignment and validation.
* **Tests**
* Added tests validating resource-group labeling and node acceptance
behavior.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: chyezh <chyezh@outlook.com>
Introduce a ScannerStartupDelay configuration to enable WAL write-only
recovery, allowing fence messages to be persisted during
primary–secondary switchover when the StreamingNode is trapped in crash
loops.
issue: https://github.com/milvus-io/milvus/issues/46368
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a configurable WAL scanner pause/resume and a consumer request
flag to optionally ignore pause signals.
* **Metrics**
* Added a scanner pause gauge and pause-duration tracking for WAL
scanning.
* **Tests**
* Added coverage for pause-consumption behavior and cleanup in stream
client tests.
* **Chores**
* Consolidated flush-all logging into a single field and added a helper
for bulk message conversion.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
issue: #46393
RO node can be created from two sources: stopping a QueryNode or replica
node transfer (e.g., suspend node). Before this fix, there were two
defects and one constraint that caused a deadlock:
Defects:
1. LeaderChecker does not sync segment distribution to RO nodes
2. Scheduler only cancels tasks on stopping nodes, not RO nodes
Constraint:
- Balance channel task blocks waiting for new delegator to become
serviceable (via sync segment) before executing release action
Deadlock scenario:
When target node becomes RO node (but not stopping) during balance
channel execution, the task gets stuck because:
- Cannot sync segment to RO node (defect 1) -> task blocks
- Task is not cancelled since node is not stopping (defect 2)
PR #45949 attempted to fix defect 1 but was not successful.
This PR unifies RO node handling by:
- LeaderChecker: only sync segment distribution to RW nodes
- Scheduler: cancel task when target node becomes RO node
- Simplify checkStale logic with unified node state checking
Signed-off-by: Wei Liu <wei.liu@zilliz.com>
Related to #46358
Refactor segment loading to use a unified diff-based approach for both
initial Load and Reopen operations:
- Extract ApplyLoadDiff from Reopen to share loading logic
- Add GetLoadDiff to compute diff from empty state for initial load
- Change column_groups_to_load from map to vector<pair> to preserve
order
- Add validation for empty index file paths in diff computation
- Add comprehensive unit tests for GetLoadDiff
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Performance**
* Improved segment loading efficiency through incremental updates,
reducing memory overhead and enhancing performance during data updates.
* **Tests**
* Expanded test coverage for load operation scenarios.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
related: #45993
This commit extends nullable vector support to the proxy layer,
querynode,
and adds comprehensive validation, search reduce, and field data
handling
for nullable vectors with sparse storage.
Proxy layer changes:
- Update validate_util.go checkAligned() with getExpectedVectorRows()
helper
to validate nullable vector field alignment using valid data count
- Update checkFloatVectorFieldData/checkSparseFloatVectorFieldData for
nullable vector validation with proper row count expectations
- Add FieldDataIdxComputer in typeutil/schema.go for logical-to-physical
index translation during search reduce operations
- Update search_reduce_util.go reduceSearchResultData to use
idxComputers
for correct field data indexing with nullable vectors
- Update task.go, task_query.go, task_upsert.go for nullable vector
handling
- Update msg_pack.go with nullable vector field data processing
QueryNode layer changes:
- Update segments/result.go for nullable vector result handling
- Update segments/search_reduce.go with nullable vector offset
translation
Storage and index changes:
- Update data_codec.go and utils.go for nullable vector serialization
- Update indexcgowrapper/dataset.go and index.go for nullable vector
indexing
Utility changes:
- Add FieldDataIdxComputer struct with Compute() method for efficient
logical-to-physical index mapping across multiple field data
- Update EstimateEntitySize() and AppendFieldData() with fieldIdxs
parameter
- Update funcutil.go with nullable vector support functions
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Full support for nullable vector fields (float, binary, float16,
bfloat16, int8, sparse) across ingest, storage, indexing, search and
retrieval; logical↔physical offset mapping preserves row semantics.
* Client: compaction control and compaction-state APIs.
* **Bug Fixes**
* Improved validation for adding vector fields (nullable + dimension
checks) and corrected search/query behavior for nullable vectors.
* **Chores**
* Persisted validity maps with indexes and on-disk formats.
* **Tests**
* Extensive new and updated end-to-end nullable-vector tests.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: marcelo-cjl <marcelo.chen@zilliz.com>
### **User description**
Issue: #46504
test: create e2e test case for highlighter
On branch feature/highlighter
Changes to be committed:
new file: milvus_client/test_milvus_client_highlighter.py
___
### **PR Type**
Tests
___
### **Description**
- Add comprehensive e2e test suite for LexicalHighlighter functionality
- Test highlighter initialization with collection setup and data
insertion
- Validate highlighter with various parameters (tags, fragments,
offsets)
- Test edge cases including Chinese characters, long text, and invalid
inputs
- Verify error handling for invalid fragment sizes, offsets, and
configurations
___
### Diagram Walkthrough
```mermaid
flowchart LR
A["Test Suite Setup"] --> B["Highlighter Init Tests"]
B --> C["Valid Test Cases"]
C --> D["Fragment Parameters"]
C --> E["Search Variations"]
C --> F["Language Support"]
B --> G["Invalid Test Cases"]
G --> H["Parameter Validation"]
G --> I["Error Handling"]
```
<details><summary><h3>File Walkthrough</h3></summary>
<table><thead><tr><th></th><th align="left">Relevant
files</th></tr></thead><tbody><tr><td><strong>Tests</strong></td><td><table>
<tr>
<td>
<details>
<summary><strong>test_milvus_client_highlighter.py</strong><dd><code>Add
comprehensive LexicalHighlighter e2e test suite</code>
</dd></summary>
<hr>
tests/python_client/milvus_client/test_milvus_client_highlighter.py
<ul><li>Create new test file with 1163 lines of comprehensive
highlighter test <br>cases<br> <li> Implement
<code>TestMilvusClientHighlighterInit</code> class to initialize
<br>collection with pre-defined test data including English, Chinese,
and <br>long text samples<br> <li> Implement
<code>TestMilvusClientHighlighterValid</code> class with 15+ test
methods <br>covering basic usage, multiple tags, fragment parameters,
offsets, <br>numbers, sentences, and language support<br> <li> Implement
<code>TestMilvusClientHighlighterInvalid</code> class with 8+ test
<br>methods validating error handling for invalid parameters and
<br>configurations<br> <li> Test highlighter with BM25 search, text
matching, and various analyzer <br>configurations</ul>
</details>
</td>
<td><a
href="https://github.com/milvus-io/milvus/pull/46505/files#diff-443e3fefb65fbdb088d5920083306ecfe3605745b1e2714198c6566ca67b3736">+1163/-0</a></td>
</tr>
</table></td></tr></tbody></table>
</details>
___
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Added a comprehensive highlighter test suite covering:
- Core highlighting with single and multi-analyzer setups and multi-tag
variations
- Fragment parameter behaviors and edge cases (size, offset, count)
- Text-match and query-based highlighting, including BM25 and vector
interactions
- Sub-word, long-text/tag, case sensitivity, Chinese/multi-language
scenarios
- Error handling for invalid parameters, no-match cases, and other edge
conditions
- Module-scoped fixture preparing multilingual, long-form test data and
teardown
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Eric Hou <eric.hou@zilliz.com>
Co-authored-by: Eric Hou <eric.hou@zilliz.com>
issue: https://github.com/milvus-io/milvus/issues/45525
see added README.md for added optimizations
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added query expression optimization feature with a new `optimizeExpr`
configuration flag to enable automatic simplification of filter
predicates, including range predicate optimization, merging of IN/NOT IN
conditions, and flattening of nested logical operators.
* **Bug Fixes**
* Adjusted delete operation behavior to correctly handle expression
evaluation.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
issue: #46274
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Performance Improvements**
* Field-level text index creation and JSON-key statistics now run
concurrently, reducing overall indexing time and speeding task
completion.
* **Observability Enhancements**
* Per-task and per-field logging expanded with richer context and
per-phase elapsed-time reporting for improved monitoring and
diagnostics.
* **Refactor**
* Node slot handling simplified to compute slot counts on demand instead
of storing them.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
relate: https://github.com/milvus-io/milvus/issues/42589
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## New Features
- Added `concurrency_per_cpu_core` configuration parameter for the
analyzer component, enabling customizable per-CPU concurrency tuning
(default: 8).
## Tests
- Added test coverage for batch analysis operations.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: aoiasd <zhicheng.yue@zilliz.com>
### **User description**
As of https://github.com/milvus-io/pymilvus/pull/2976 `milvus-lite` is
no longer included as part of the default `pymilvus` installation and
must be explicitly specified
___
### **PR Type**
Documentation
___
### **Description**
- Update installation docs to reflect milvus-lite as optional dependency
- Clarify explicit installation requirement for Milvus Lite
functionality
___
### Diagram Walkthrough
```mermaid
flowchart LR
A["pymilvus default install"] -- "does not include" --> B["Milvus Lite"]
C["pymilvus[milvus-lite]"] -- "includes" --> B
B -- "enables" --> D["Local vector database"]
```
<details><summary><h3>File Walkthrough</h3></summary>
<table><thead><tr><th></th><th align="left">Relevant
files</th></tr></thead><tbody><tr><td><strong>Documentation</strong></td><td><table>
<tr>
<td>
<details>
<summary><strong>README.md</strong><dd><code>Clarify Milvus Lite as
optional dependency</code>
</dd></summary>
<hr>
README.md
<ul><li>Updated installation documentation to clarify that
<code>milvus-lite</code> is no <br>longer included by default<br> <li>
Changed wording from "includes Milvus Lite" to "can try Milvus Lite by
<br>installing <code>pymilvus[milvus-lite]</code>"<br> <li> Reflects the
breaking change from pymilvus PR #2976 where milvus-lite <br>became an
optional dependency</ul>
</details>
</td>
<td><a
href="https://github.com/milvus-io/milvus/pull/46474/files#diff-b335630551682c19a781afebcf4d07bf978fb1f8ac04c6bf87428ed5106870f5">+1/-1</a>
</td>
</tr>
</table></td></tr></tbody></table>
</details>
___
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Documentation**
* Updated Milvus Lite installation guidance to specify the proper
installation method using the optional feature specification.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Nathan Weinberg <nathan2@stwmd.net>
issue: #45486
Introduce row group batching to reduce cache cell granularity and
improve
memory&disk efficiency. Previously, each parquet row group mapped 1:1 to
a cache
cell. Now, up to `kRowGroupsPerCell` (4) row groups are merged into one
cell.
This reduces the number of cache cells (and associated overhead) by ~4x
while
maintaining the same data granularity for loading.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Refactor**
* Switched to cell-based grouping that merges multiple row groups for
more efficient multi-file aggregation and reads.
* Chunk loading now combines multiple source batches/tables per cell and
better supports mmap-backed storage.
* **New Features**
* Exposed helpers to query row-group ranges and global row-group offsets
for diagnostics and testing.
* Translators now accept chunk-type and mmap/load hints to control
on-disk vs in-memory behavior.
* **Bug Fixes**
* Improved bounds checks and clearer error messages for out-of-range
cell requests.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Shawn Wang <shawn.wang@zilliz.com>
### **User description**
Related to #44956
Add manifest-based data loading path for optional fields in
`cache_opt_field_memory_v2`. When a manifest file is provided in the
config, the function now retrieves field data directly from the manifest
using `GetFieldDatasFromManifest` instead of reading from segment insert
files. This enables storage v2 compatibility for building indexes with
optional fields.
___
### **PR Type**
Enhancement
___
### **Description**
- Add manifest-based data loading for optional fields in index building
- Support storage v2 compatibility via `GetFieldDatasFromManifest`
function
- Enable PK isolation optional field handling without segment insert
files
___
---------
Signed-off-by: Congqi Xia <congqi.xia@zilliz.com>
issue: https://github.com/milvus-io/milvus/issues/44399
this PR also adds `ByteSize()` methods for scalar indexes. currently not
used in milvus code, but used in scalar benchmark. may be used by
cachinglayer in the future.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Refactor**
* Improved and standardized memory-size computation and caching across
index types so reported index footprints are more accurate and
consistent.
* **Chores**
* Ensured byte-size metrics are refreshed immediately after index
build/load operations to keep memory accounting in sync with runtime
state.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
This commit refines L0 compaction to ensure data consistency by properly
setting the delete position boundary for L1 segment selection.
Key Changes:
1. L0 View Trigger Sets latestDeletePos for L1 Selection
2. Filter L0 Segments by Growing Segment Position in policy, not in
views
3. Renamed LevelZeroSegmentsView to LevelZeroCompactionView
4. Renamed fields for semantic clarity: * segments -> l0Segments *
earliestGrowingSegmentPos -> latestDeletePos
5. Update Default Compaction Prioritizer to level
See also: #46434
---------
Signed-off-by: yangxuan <xuan.yang@zilliz.com>
This PR improves the robustness of object storage operations by retrying
both explicit throttling errors (e.g. HTTP 429, SlowDown, ServerBusy).
These errors commonly occur under high concurrency and are typically
recoverable with bounded retries.
issue: https://github.com/milvus-io/milvus/issues/44772
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Configurable retry support for reads from object storage and improved
mapping of transient/rate-limit errors.
* Added a retryable reader wrapper used by CSV/JSON/Parquet/Numpy import
paths.
* **Configuration**
* New parameter to control storage read retry attempts.
* **Tests**
* Expanded unit tests covering error mapping and retry behaviors across
storage backends.
* Standardized mock readers and test initialization to simplify test
setups.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
issue: #46349
When using brute-force search, the iterator results from multiple chunks
are merged; at that point, we need to pay attention to how the metric
affects result ranking.
Signed-off-by: xianliang.li <xianliang.li@zilliz.com>
generated a library that wraps the go expr parser, and embedded that
into libmilvus-core.so
issue: https://github.com/milvus-io/milvus/issues/45702
see `internal/core/src/plan/milvus_plan_parser.h` for the exposed
interface
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Introduced C++ API for plan parsing with schema registration and
expression parsing capabilities.
* Plan parser now available as shared libraries instead of a standalone
binary tool.
* **Refactor**
* Reorganized build system to produce shared library artifacts instead
of executable binaries.
* Build outputs relocated to standardized library and include
directories.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: Buqian Zheng <zhengbuqian@gmail.com>
### **User description**
issue: #46507
we use the assign/unassign api to manage the consumer manually, the
commit operation will generate a new consumer group which is not what we
want. so we disable the auto commit to avoid it, also see:
https://github.com/confluentinc/confluent-kafka-python/issues/250#issuecomment-331377925
___
### **PR Type**
Bug fix
___
### **Description**
- Disable auto-commit in Kafka consumer configuration
- Prevents unwanted consumer group creation from manual offset
management
- Clarifies offset reset behavior with explanatory comments
___
### Diagram Walkthrough
```mermaid
flowchart LR
A["Kafka Consumer Config"] --> B["Set enable.auto.commit to false"]
B --> C["Prevent auto consumer group creation"]
A --> D["Set auto.offset.reset to earliest"]
D --> E["Handle deleted offsets gracefully"]
```
<details><summary><h3>File Walkthrough</h3></summary>
<table><thead><tr><th></th><th align="left">Relevant
files</th></tr></thead><tbody><tr><td><strong>Bug
fix</strong></td><td><table>
<tr>
<td>
<details>
<summary><strong>builder.go</strong><dd><code>Disable auto-commit and
add configuration comments</code>
</dd></summary>
<hr>
pkg/streaming/walimpls/impls/kafka/builder.go
<ul><li>Added <code>enable.auto.commit</code> configuration set to
<code>false</code> to prevent <br>automatic consumer group creation<br>
<li> Added explanatory comments for both <code>auto.offset.reset</code>
and <br><code>enable.auto.commit</code> settings<br> <li> Clarifies that
manual assign/unassign API is used for consumer <br>management</ul>
</details>
</td>
<td><a
href="https://github.com/milvus-io/milvus/pull/46508/files#diff-4b5635821fdc8b585d16c02d8a3b59079d8e667b2be43a073265112d72701add">+7/-0</a>
</td>
</tr>
</table></td></tr></tbody></table>
</details>
___
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Bug Fixes
* Kafka consumer now reads from the earliest available messages and
auto-commit has been disabled to support manual offset management.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: chyezh <chyezh@outlook.com>
### **PR Type**
Enhancement
___
### **Description**
- Only flush and fence segments for schema-changing alter collection
messages
- Skip segment sealing for collection property-only alterations
- Add conditional check using messageutil.IsSchemaChange utility
function
___
### Diagram Walkthrough
```mermaid
flowchart LR
A["Alter Collection Message"] --> B{"Is Schema Change?"}
B -->|Yes| C["Flush and Fence Segments"]
B -->|No| D["Skip Segment Operations"]
C --> E["Set Flushed Segment IDs"]
D --> E
E --> F["Append Operation"]
```
<details><summary><h3>File Walkthrough</h3></summary>
<table><thead><tr><th></th><th align="left">Relevant
files</th></tr></thead><tbody><tr><td><strong>Enhancement</strong></td><td><table>
<tr>
<td>
<details>
<summary><strong>shard_interceptor.go</strong><dd><code>Conditional
segment sealing based on schema changes</code>
</dd></summary>
<hr>
internal/streamingnode/server/wal/interceptors/shard/shard_interceptor.go
<ul><li>Added import for <code>messageutil</code> package to access
schema change detection <br>utility<br> <li> Modified
<code>handleAlterCollection</code> to conditionally flush and fence
<br>segments only for schema-changing messages<br> <li> Wrapped segment
flushing logic in <code>if
</code><br><code>messageutil.IsSchemaChange(header)</code> check<br>
<li> Skips unnecessary segment sealing when only collection properties
are <br>altered</ul>
</details>
</td>
<td><a
href="https://github.com/milvus-io/milvus/pull/46488/files#diff-c1acf785e5b530e59137b21584cf567ccd9aeeb613fb3684294b439289e80beb">+9/-3</a>
</td>
</tr>
</table></td></tr></tbody></table>
</details>
___
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Optimized collection schema alteration to conditionally perform
segment allocation operations only when schema changes are detected,
reducing unnecessary overhead in unmodified collection scenarios.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
issue: #46451
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Session versioning added to validate coordinator compatibility during
registration and active takeover.
* **Changes**
* Active–standby flow simplified: standby-to-active activation now
always enabled and initialized unconditionally.
* Registration uses version-aware transactions to ensure version
consistency during takeover.
* Startup/health startup path streamlined.
* **Tests**
* Added version-key integration test; removed test for disabling
active-standby.
* Updated flush test to assert rate-limiter errors occur.
* **Chores**
* Removed centralized connection manager and its test suite.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Signed-off-by: chyezh <chyezh@outlook.com>
### **User description**
issue: #46097
- the flush rate is modified into 4qps, so the testcase is fail.
___
### **PR Type**
Tests, Bug fix
___
### **Description**
- Replace sequential flush calls with concurrent requests to trigger
rate limiting
- Add sync.WaitGroup for concurrent goroutine execution
- Check for rate limit errors across multiple concurrent flush
operations
- Remove hardcoded error message expectation for flexibility
___
### Diagram Walkthrough
```mermaid
flowchart LR
A["Sequential Flush Calls"] -->|Replace with| B["Concurrent Flush Requests"]
B -->|Use| C["sync.WaitGroup"]
C -->|Validate| D["Rate Limit Errors"]
```
<details><summary><h3>File Walkthrough</h3></summary>
<table><thead><tr><th></th><th align="left">Relevant
files</th></tr></thead><tbody><tr><td><strong>Tests</strong></td><td><table>
<tr>
<td>
<details>
<summary><strong>insert_test.go</strong><dd><code>Refactor flush rate
test to use concurrent requests</code>
</dd></summary>
<hr>
tests/go_client/testcases/insert_test.go
<ul><li>Added <code>sync</code> package import for concurrent goroutine
synchronization<br> <li> Replaced sequential flush calls with 10
concurrent flush operations <br>using goroutines<br> <li> Implemented
WaitGroup to synchronize all concurrent flush requests<br> <li> Modified
error validation to check for rate limit errors across all
<br>concurrent attempts instead of expecting specific sequential
behavior<br> <li> Relaxed error message matching to only check for "rate
limit exceeded" <br>substring</ul>
</details>
</td>
<td><a
href="https://github.com/milvus-io/milvus/pull/46497/files#diff-89a4ddfa15d096e6a5f647da0e461715e5a692b375b04a3d01939f419b00f529">+19/-4</a>
</td>
</tr>
</table></td></tr></tbody></table>
</details>
___
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Release Notes
* **Tests**
* Enhanced testing of concurrent flush operations to improve validation
of system reliability under concurrent load scenarios.
---
**Note:** This release contains internal testing improvements with no
direct user-facing feature changes.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Signed-off-by: chyezh <chyezh@outlook.com>
AddProxyClients now removes clients not in the new snapshot before
adding new ones. This ensures proper cleanup when ProxyWatcher re-watche
etcd.
issue: https://github.com/milvus-io/milvus/issues/46397
Signed-off-by: bigsheeper <yihao.dai@zilliz.com>
