feat: Add MLflow tracing, trace Assessment Sync, and Trace Export by Vperiodt · Pull Request #6405 · feast-dev/feast

Vperiodt · 2026-05-14T13:34:53Z

What this PR does / why we need it

Adds end-to-end MLflow integration for Feast: distributed tracing, trace export, assessment sync, and MLflow GenAI Dataset sync — enabling ML teams to trace agent-feature interactions, produce training data from production traces, and serve curated evaluation datasets through Feast's online/offline stores.

1. Distributed Tracing (`mlflow.enable_distributed_tracing: true`)

Server-side spans: Every feature retrieval endpoint (get-online-features, search, write-to-online-store) creates an MLflow span with feature names, entity count, and project metadata.
Cross-process trace linking: W3C traceparent header propagation — Feast spans become children of the caller's trace. MCP server instruments its internal httpx client and forwards mcp-session-id for session grouping.
In-process LLM span tagging: A span processor (feast_span_processor) automatically stamps LLM/CHAT_MODEL spans with feast.context_features when the SDK is used directly via feast_trace_scope().
PII redaction: Optional redact_entity_pii scrubs entity values from span inputs.
Sampling: trace_sampling_ratio controls what fraction of requests produce spans.
Graceful degradation: Missing MLflow, incompatible versions, or span creation failures never block feature serving.

2. Trace Assessment Sync (`feast mlflow sync-assessments`)

Assessment ingestion: Scans MLflow traces in an experiment for expectations and feedback, writes them to a Feast FeatureView or LabelView.
Flat mode (default): Each assessment = one row with trace_id, assessment_name, assessment_type, value, source_id, rationale, event_timestamp.
Pivot mode (--pivot): Groups assessments by trace_id into LabelView-compatible rows with --assessment-mapping to control column names and --labeler-column for the assessment source.
Selective sync: --assessment-names filters which assessments to ingest.
LabelView support: Falls back to store.get_label_view() when store.get_feature_view() fails, enabling correct offline writes via PushSource.

3. Trace Export (`feast mlflow export-traces`)

Trace extraction: Queries MLflow for CHAT_MODEL spans, extracts prompt/completion pairs with feast.* context attributes and assessments (expectations + feedback).
Label resolution: Two paths — join with a Feast LabelView (offline ConflictPolicy or online LAST_WRITE_WINS) or promote MLflow expectations to corrected_response.
Dataset filtering: --dataset limits export to traces curated into a named MLflow GenAI Dataset.
Export formats: --format openai (OpenAI-compatible JSONL) or --format enriched (includes trace provenance, feature refs, entity values, label info).
Artifact registration: --register saves exported JSONL as an MLflow artifact; --register-experiment / --register-run control where.

Configuration:
   mlflow:
  enabled: true
  tracking_uri: "http://mlflow:5000"
  enable_distributed_tracing: true
  trace_sampling_ratio: 1.0
  redact_entity_pii: false

Which issue(s) this PR fixes:

Checks

I've made sure the tests are passing.
My commits are signed off (git commit -s)
My PR title follows conventional commits format

Testing Strategy

Unit tests
Integration tests
Manual tests
Testing is not required for this change

Misc

devin-ai-integration

Devin Review found 1 new potential issue.

View 7 additional findings in Devin Review.

devin-ai-integration · 2026-05-27T14:10:37Z

🔴 Removing detect_types from SQLite connection breaks event_ts in retrieve_online_documents_v2

The PR removes detect_types=sqlite3.PARSE_DECLTYPES | sqlite3.PARSE_COLNAMES from _initialize_conn at sqlite.py:705. This flag was responsible for automatically converting stored timestamps back to datetime objects via the registered converters (lines 99-101). Since datetime objects are stored as epoch integers (via the adapt_datetime_epoch adapter at line 81), removing detect_types means timestamp columns now return raw int values instead of datetime objects.

The online_read method was correctly updated (lines 278-285) to handle int timestamps. However, retrieve_online_documents_v2 was NOT updated. At line 677, the check isinstance(entity_dict[entity_key_value]["event_ts"], datetime) will always be False because event_ts is now an int, causing res_event_ts to always be None. This loses all timestamp information for document retrieval results, which flows into _retrieve_from_online_store_v2 (feature_store.py:3575) and ultimately into the API response.

(Refers to lines 677-678)
$Open in Devin Review$

Was this helpful? React with 👍 or 👎 to provide feedback.

franciscojavierarceo · 2026-05-27T15:24:36Z

linter is failing

ntkathole · 2026-06-08T12:54:46Z

@@ -0,0 +1,130 @@
+# Feast MLflow Tracing


Add this official docs

jyejare

This PR adds comprehensive MLflow distributed tracing support to Feast, enabling automatic trace collection for feature server endpoints, cross-process trace linking via traceparent headers, and in-process feature context tagging. The implementation is well-structured with proper separation of concerns, comprehensive test coverage, and graceful degradation when dependencies are missing.

jyejare · 2026-06-09T18:12:42Z

+                if isinstance(ts, (int, float)):
+                    res_ts = datetime.fromtimestamp(ts, tz=timezone.utc)
                else:
-                    res_ts = ts.replace(tzinfo=timezone.utc)
+                    ts = cast(datetime, ts)
+                    if ts.tzinfo is not None:
+                        res_ts = ts.astimezone(timezone.utc)
+                    else:


[Warning] Inconsistent timestamp handling may cause runtime errors

The code handles both int/float and datetime timestamp types, but the original code assumed datetime objects. This change could break existing code that expects consistent datetime handling and may cause timezone issues with numeric timestamps.

Suggested:

Suggested change

if isinstance(ts, (int, float)):

res_ts = datetime.fromtimestamp(ts, tz=timezone.utc)

else:

res_ts = ts.replace(tzinfo=timezone.utc)

ts = cast(datetime, ts)

if ts.tzinfo is not None:

res_ts = ts.astimezone(timezone.utc)

else:

ts = cast(datetime, ts)

if ts.tzinfo is not None:

res_ts = ts.astimezone(timezone.utc)

else:

res_ts = ts.replace(tzinfo=timezone.utc)

jyejare · 2026-06-09T18:12:42Z

 oracle = ["ibis-framework[oracle]>=10.0.0"]
 mysql = ["pymysql", "types-PyMySQL"]
 openlineage = ["openlineage-python>=1.40.0"]
+mlflow = [
+    "mlflow>=2.14.0",
+    "opentelemetry-api>=1.28.0",
+    "opentelemetry-sdk>=1.28.0",
+    "opentelemetry-instrumentation-fastapi>=0.49b0",
+    "opentelemetry-instrumentation-httpx>=0.49b0",


[Suggestion] Consider adding opentelemetry-exporter-otlp-proto-http dependency

The mlflow extra includes several OpenTelemetry packages but is missing opentelemetry-exporter-otlp-proto-http which was mentioned in the pixi.lock file. This could cause issues if OTLP export is needed.

Suggested:

Suggested change

oracle = ["ibis-framework[oracle]>=10.0.0"]

mysql = ["pymysql", "types-PyMySQL"]

openlineage = ["openlineage-python>=1.40.0"]

mlflow = [

"mlflow>=2.14.0",

"opentelemetry-api>=1.28.0",

"opentelemetry-sdk>=1.28.0",

"opentelemetry-instrumentation-fastapi>=0.49b0",

"opentelemetry-instrumentation-httpx>=0.49b0",

mlflow = [

"mlflow>=2.14.0",

"opentelemetry-api>=1.28.0",

"opentelemetry-sdk>=1.28.0",

"opentelemetry-instrumentation-fastapi>=0.49b0",

"opentelemetry-instrumentation-httpx>=0.49b0",

"opentelemetry-exporter-otlp-proto-http>=1.28.0",

]

Added that before it was causing grpcio conflicts and
opentelemetry-exporter-otlp-proto-http is already a transitive dep of mlflow>=2.14.0

jyejare · 2026-06-09T18:12:42Z

+        return
+
+    try:
+        from opentelemetry.instrumentation.fastapi import FastAPIInstrumentor
+
+        FastAPIInstrumentor.instrument_app(app)
+        logger.info("FastAPI OTEL instrumentation enabled for trace propagation")
+    except ImportError:
+        logger.debug(
+            "opentelemetry-instrumentation-fastapi not installed; "
+            "cross-process trace linking disabled"
+        )
+


[Suggestion] Improve error handling in FastAPI instrumentation

The ImportError handling is good, but it would be helpful to also catch other potential exceptions during instrumentation and provide more specific error messages.

Suggested:

Suggested change

return

try:

from opentelemetry.instrumentation.fastapi import FastAPIInstrumentor

FastAPIInstrumentor.instrument_app(app)

logger.info("FastAPI OTEL instrumentation enabled for trace propagation")

except ImportError:

logger.debug(

"opentelemetry-instrumentation-fastapi not installed; "

"cross-process trace linking disabled"

)

try:

from opentelemetry.instrumentation.fastapi import FastAPIInstrumentor

FastAPIInstrumentor.instrument_app(app)

logger.info("FastAPI OTEL instrumentation enabled for trace propagation")

except ImportError:

logger.debug(

"opentelemetry-instrumentation-fastapi not installed; "

"cross-process trace linking disabled"

)

except Exception as e:

logger.warning(f"Failed to instrument FastAPI app: {e}")

jyejare · 2026-06-09T18:12:42Z

+        # async exporter must send the span before that happens.
+        if has_traceparent:
+            try:
+                _mlflow_mod.flush_trace_async_logging()
+            except Exception:


[Warning] Potential performance impact from synchronous flush in async context

Calling flush_trace_async_logging() synchronously in an async endpoint could block the event loop and impact performance. Consider making this operation truly asynchronous or moving it to a background task.

jyejare · 2026-06-09T18:12:42Z

+def feast_trace_scope() -> Iterator[FeastTraceContext]:
+    """Context manager that creates and cleans up a ``FeastTraceContext``."""
+    ctx = FeastTraceContext()
+    _thread_local.feast_ctx = ctx
+    try:
+        yield ctx
+    finally:
+        ctx.clear()


[Suggestion] Consider adding context validation

The feast_trace_scope context manager could benefit from validation to ensure the context is properly cleaned up even if exceptions occur during setup.

Suggested:

Suggested change

def feast_trace_scope() -> Iterator[FeastTraceContext]:

"""Context manager that creates and cleans up a ``FeastTraceContext``."""

ctx = FeastTraceContext()

_thread_local.feast_ctx = ctx

try:

yield ctx

finally:

ctx.clear()

@contextmanager

def feast_trace_scope() -> Iterator[FeastTraceContext]:

"""Context manager that creates and cleans up a ``FeastTraceContext``."""

old_ctx = getattr(_thread_local, 'feast_ctx', None)

ctx = FeastTraceContext()

_thread_local.feast_ctx = ctx

try:

yield ctx

finally:

ctx.clear()

_thread_local.feast_ctx = old_ctx

jyejare · 2026-06-09T18:12:42Z

+    enable_tracing: StrictBool = True
+    """ bool: When True and mlflow.enabled=True, server-side API calls
+        create MLflow trace spans via mlflow.start_span().  Spans appear
+        in the MLflow UI Traces tab and support parent-child linking via
+        traceparent headers.  Defaults to True. """
+


[Nitpick] Consider more descriptive field name

The field name 'enable_tracing' could be more specific since it only controls MLflow tracing, not all tracing in Feast.

Suggested:

Suggested change

enable_tracing: StrictBool = True

""" bool: When True and mlflow.enabled=True, server-side API calls

create MLflow trace spans via mlflow.start_span(). Spans appear

in the MLflow UI Traces tab and support parent-child linking via

traceparent headers. Defaults to True. """

enable_distributed_tracing: StrictBool = True

""" bool: When True and mlflow.enabled=True, server-side API calls

create MLflow trace spans via mlflow.start_span(). Spans appear

in the MLflow UI Traces tab and support parent-child linking via

traceparent headers. Defaults to True. """

ntkathole · 2026-06-11T06:31:31Z

@@ -0,0 +1,132 @@
+# MLflow Distributed Tracing


better to include this in existing mlflow.md

ntkathole · 2026-06-11T06:32:10Z

 * [\[Alpha\] Feature View Versioning](reference/alpha-feature-view-versioning.md)
 * [OpenLineage Integration](reference/openlineage.md)
 * [MLflow Integration](reference/mlflow.md)
+* [MLflow Distributed Tracing](reference/mlflow-distributed-tracing.md)


no need of different section

ntkathole · 2026-06-11T06:38:12Z

+            _logger.debug("Failed to set attribute %s on LLM span", key)
+
+
+def install_feast_span_processor() -> None:


is it being called somewhere other than tests ?

It is expected to be called by the users in there agent code which helps to tag LLM spans with feast context, either i can document it or we can add it in the _lazy_init()

I updated it in the docs

ntkathole · 2026-07-19T11:12:07Z

@@ -0,0 +1,222 @@
+"""CLI commands for fine-tuning dataset export."""


feast finetuning is not correct naming here.

The feast finetuning name implies Feast has an opinion about LLM training workflows, which is a much larger product decision than what's in this PR.

I'ts better if command is like feast mlflow export-traces or similar - making it clear this is an MLflow integration subcommand

ntkathole · 2026-07-19T11:14:57Z

+    if dry_run:
+        click.echo("  DRY RUN — no data will be written.")
+
+    result = sync_mlflow_dataset_to_feast(


I think better integration will be if it's MlflowDatasetSource. The dataset sync should be an MlflowDatasetSource (proper Feast primitive).

codecov-commenter · 2026-07-21T11:40:11Z

⚠️ Please install the $'codecov app svg image'$ to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 42.47076% with 787 lines in your changes missing coverage. Please review.
✅ Project coverage is 46.75%. Comparing base (037c4cd) to head (be79ec5).

Files with missing lines	Patch %	Lines
sdk/python/feast/trace_export/trace_extractor.py	29.72%	171 Missing and 11 partials ⚠️
...dk/python/feast/mlflow_integration/dataset_sync.py	37.71%	124 Missing and 23 partials ⚠️
sdk/python/feast/feature_server.py	32.55%	115 Missing and 1 partial ⚠️
sdk/python/feast/trace_export/label_resolver.py	25.58%	94 Missing and 2 partials ⚠️
sdk/python/feast/cli/mlflow_cmd.py	30.82%	92 Missing ⚠️
sdk/python/feast/trace_export/dataset_filter.py	0.00%	45 Missing ⚠️
sdk/python/feast/tracing.py	72.80%	27 Missing and 7 partials ⚠️
sdk/python/feast/trace_export/exporters.py	74.41%	14 Missing and 8 partials ⚠️
sdk/python/feast/mlflow_integration/__init__.py	13.33%	13 Missing ⚠️
sdk/python/feast/trace_export/__init__.py	18.75%	13 Missing ⚠️
... and 5 more
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

$Impacted file tree graph$

@@            Coverage Diff             @@
##           master    #6405      +/-   ##
==========================================
- Coverage   46.76%   46.75%   -0.01%     
==========================================
  Files         415      425      +10     
  Lines       50392    51706    +1314     
  Branches     7214     7480     +266     
==========================================
+ Hits        23567    24177     +610     
- Misses      25171    25820     +649     
- Partials     1654     1709      +55

Flag	Coverage Δ
go-feature-server	`30.58% <ø> (ø)`
python-unit	`48.04% <42.47%> (-0.05%)`	⬇️

Files with missing lines	Coverage Δ
sdk/python/feast/cli/cli.py	`55.51% <100.00%> (+0.34%)`	⬆️
sdk/python/feast/mlflow_integration/config.py	`91.83% <100.00%> (+91.83%)`	⬆️
sdk/python/feast/tracing_context.py	`95.91% <95.91%> (ø)`
sdk/python/feast/tracing_hooks.py	`90.00% <90.00%> (ø)`
sdk/python/feast/infra/mcp_servers/mcp_server.py	`81.70% <66.66%> (-3.59%)`	⬇️
sdk/python/feast/feature_store.py	`43.03% <52.94%> (+0.06%)`	⬆️
sdk/python/feast/infra/online_stores/sqlite.py	`59.33% <27.27%> (-0.97%)`	⬇️
sdk/python/feast/mlflow_integration/__init__.py	`31.57% <13.33%> (+31.57%)`	⬆️
sdk/python/feast/trace_export/__init__.py	`18.75% <18.75%> (ø)`
sdk/python/feast/trace_export/exporters.py	`74.41% <74.41%> (ø)`
... and 7 more

... and 2 files with indirect coverage changes

Continue to review full report in Codecov by Harness.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 037c4cd...be79ec5. Read the comment docs.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

ntkathole · 2026-07-29T12:33:35Z

@Vperiodt Please rebase and solve conflicts, also please fix the linting

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com>

$@Vperiodt$ Vperiodt changed the title ~~mlflow-tracing~~ [WIP] mlflow-tracing May 14, 2026

$@Vperiodt$ Vperiodt changed the title ~~[WIP] mlflow-tracing~~ [WIP] : mlflow-tracing May 24, 2026

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from bfc5f92 to e89e0a6 Compare May 25, 2026 06:29

$@Vperiodt$ Vperiodt changed the title ~~[WIP] : mlflow-tracing~~ feat: MLflow native distributed tracing for Feast feature server May 26, 2026

$@Vperiodt$

Vperiodt marked this pull request as ready for review May 26, 2026 14:59

$@Vperiodt$

Vperiodt requested a review from a team as a code owner May 26, 2026 14:59

$devin-ai-integration[bot]$

This comment was marked as resolved.

Sign in to view

$@Vperiodt$ Vperiodt changed the title ~~feat: MLflow native distributed tracing for Feast feature server~~ feat: MLflow distributed tracing for Feast feature server May 27, 2026

$devin-ai-integration[bot]$

devin-ai-integration Bot reviewed May 27, 2026

View reviewed changes

$ntkathole$

ntkathole reviewed Jun 8, 2026

View reviewed changes

$jyejare$

jyejare reviewed Jun 9, 2026

View reviewed changes

$ntkathole$

ntkathole reviewed Jun 11, 2026

View reviewed changes

$@Vperiodt$ Vperiodt changed the title ~~feat: MLflow distributed tracing for Feast feature server~~ feat: Add MLflow tracing, fine-tuning export, and GenAI dataset sync Jul 13, 2026

$ntkathole$

ntkathole reviewed Jul 19, 2026

View reviewed changes

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from ebd53a9 to cce1fec Compare July 24, 2026 08:51

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from d3f6565 to 0b4b400 Compare July 30, 2026 09:29

Vperiodt added 3 commits July 31, 2026 15:36

$@Vperiodt$

feat: MLflow distributed tracing for Feast feature server

b31ad85

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

$@Vperiodt$

update package name

e35f5f8

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

$@Vperiodt$

trigger CI

ed6949d

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from ee0ae39 to ed6949d Compare July 31, 2026 10:08

$@Vperiodt$

update lock-files

7228fb8

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from f75cb32 to 7228fb8 Compare July 31, 2026 11:14

$@Vperiodt$

chore: Strip DataSource code from tracing PR

855aeb9

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from 7f46aa2 to 855aeb9 Compare August 4, 2026 18:35

$@Vperiodt$

Merge branch 'master' into mcp-trace

1e34d7c

$@Vperiodt$ Vperiodt changed the title ~~feat: Add MLflow tracing, fine-tuning export, and GenAI dataset sync~~ feat: Add MLflow tracing, trace Assessment Sync, and Trace Export Aug 4, 2026

$@Vperiodt$ Vperiodt changed the title ~~feat: Add MLflow tracing, trace Assessment Sync, and Trace Export~~ feat:Add MLflow tracing, trace Assessment Sync, and Trace Export Aug 5, 2026

$@Vperiodt$ Vperiodt changed the title ~~feat:Add MLflow tracing, trace Assessment Sync, and Trace Export~~ feat: Add MLflow tracing, trace Assessment Sync, and Trace Export Aug 5, 2026

Vperiodt added 2 commits August 5, 2026 13:57

$@Vperiodt$

fix-CI

db078ab

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED

$@Vperiodt$

Merge branch 'master' into mcp-trace

513d331

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from 09e2fee to fef300c Compare August 10, 2026 08:33

$@Vperiodt$

remove protos

be79ec5

Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com> rh-pre-commit.version: 2.3.2 rh-pre-commit.check-secrets: ENABLED Signed-off-by: Vanshika Vanshika <vvanshik@redhat.com>

$@Vperiodt$

Vperiodt force-pushed the mcp-trace branch from fef300c to be79ec5 Compare August 10, 2026 09:35

		_logger.debug("Failed to set attribute %s on LLM span", key)


		def install_feast_span_processor() -> None:

		@@ -0,0 +1,222 @@
		"""CLI commands for fine-tuning dataset export."""

Uh oh!

Conversation

Vperiodt commented May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What this PR does / why we need it

1. Distributed Tracing (mlflow.enable_distributed_tracing: true)

2. Trace Assessment Sync (feast mlflow sync-assessments)

3. Trace Export (feast mlflow export-traces)

Which issue(s) this PR fixes:

Checks

Testing Strategy

Misc

Uh oh!

This comment was marked as resolved.

Uh oh!

devin-ai-integration Bot left a comment

Choose a reason for hiding this comment

Uh oh!

devin-ai-integration Bot May 27, 2026

Choose a reason for hiding this comment

Uh oh!

franciscojavierarceo commented May 27, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jyejare left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codecov-commenter commented Jul 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

ntkathole commented Jul 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

$@Vperiodt$ Vperiodt commented May 14, 2026 •

edited

Loading

1. Distributed Tracing (`mlflow.enable_distributed_tracing: true`)

2. Trace Assessment Sync (`feast mlflow sync-assessments`)

3. Trace Export (`feast mlflow export-traces`)

$@devin-ai-integration$ devin-ai-integration Bot left a comment

$@devin-ai-integration$ devin-ai-integration Bot May 27, 2026

$@jyejare$ jyejare left a comment

codecov-commenter commented Jul 21, 2026 •

edited

Loading