-
Notifications
You must be signed in to change notification settings - Fork 241
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Model Monitoring] Implement TSDB abstraction #5210
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM.
Added few more comments :)
""" | ||
Write a single application or metric to TSDB. | ||
|
||
:raise mlrun.errors.MLRunInvalidArgumentError: If an error occurred while writing the event. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Are you sure this is the right exception to raise? Sounds more like an MLRunRuntimeError
.
:param end: The end time of the metrics. | ||
|
||
:return: DataFrame with the provided attributes from the data collection. | ||
:raise: MLRunInvalidArgumentError if the provided table wasn't found. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This one sounds more like an MLRunNotFoundError
.
…the streaming interface (#5599) * [Datastore] Rename `DatastoreProfileAzureBlob` 'bucket' parameter to 'container' (#5521) * [Project] Add delete artifact method (#5491) * [API] Add delete artifact data endpoints (#5477) * [Docs] Add registering an HDFS datastore profile (#5396) To * how to register a datastore profile * Update docs/store/datastore.md Co-authored-by: Gal Topper <gal.topper@gmail.com> * Update docs/store/datastore.md Co-authored-by: Gal Topper <gal.topper@gmail.com> * fix: data store is 2 words * clarify example --------- Co-authored-by: Gal Topper <gal.topper@gmail.com> * [Alerts] Add backwards compatibility for igz < 3.6 (#5524) * [Coding Conventions] Fix `mlconf` usages (#5525) * [API] Enhance usage of iguazio client (#5506) * [Model Monitoring] Implement TSDB abstraction (#5210) * [Automation] Don't rollout deployment before purging DB (#5527) * [SQLDB] Refactor partition querying to minimize subquery table size (#5526) * [Docs] Enable automodule for mlrun.comon.schemas.artifact (#5529) * [Alerts] Filter project resources by permissions for list (#5532) * [ProjectSummaries] Add pipelines_completed_recent_count, pipelines_failed_recent_count (#5523) * [KFP] Merge KFP2 Feature Branch (#5533) * [Nuclio] Delete remote function config map (#5530) * [API] Add read model monitoring metrics values endpoint (#5519) * [Model Monitoring] Fix up imports (#5531) * Remove unused import * Change import so that PyCharm detects inheritance correctly * [Notifications] Display Step Kind In Pipeline Notification (#5538) * Add step kind to pipeline notification * better system test * redundant comment * fix ut --------- Co-authored-by: quaark <a.melnick@icloud.com> * [Alerts] Modify alert entity to have list of ids per alert (#5511) * [Application] Fix test function handler assertion (#5548) * [Docs] Feature-store + Datastore: consolidate Apache Kafka details, cleanups in datastores (#5305) * [Model Monitoring] Fix invalid current_stats in `MonitoringApplicationContext` (#5552) fix deepcopy bug * [KFP] Bump version and create makefile for release (#5555) * [Runs] Store artifact URIs instead of full artifacts in body (#5553) * [Artifacts] Add `producer_uri` filter to list artifacts API (#5549) * [API] Check the artifact kind while attempting to remove artifact data (#5545) * [CI] Fix pipeline adapters makefile (#5557) * [Model Monitoring] Update model endpoint's current stats and drift measures in the writer (#5554) * [Model Monitoring] Add new TSDB target for all predictions (#5551) * [Model Monitoring] Add new TSDB target for all predictions Tracking count and latency. [ML-6349](https://iguazio.atlassian.net/browse/ML-6349) * Change TSDB predictions table path * Move predictions table * Fix * Fix parameter names * Fix * [Functions] Update function with deletion task id (#5536) * [Application] Create API Gateway by default (#5515) * [Logger] Enhance formatter creation (#5561) * [KFP] Suppress reusable-components warnings (#5501) * [ProjectSummaries] Add distinct_scheduled_jobs_pending_count, distinct_scheduled_pipelines_pending_count (#5535) * [Notifications] Fix Pipeline Notification Log Spam + Git Notification (#5563) fix pipeline notification bugs Co-authored-by: quaark <a.melnick@icloud.com> * [Docs] Compile notification when compiling docs (#5564) * [Helpers] Ignore pre-release in `min_iguazio_version` (#5558) * [Runs] Enrich run with artifacts when getting a single run (#5565) * [Application] Set api gateway name on status (#5567) * [CI] Bundling demos as release asset (#5550) * [Nuclio-jupyter] Bump nuclio-jupyter to 0.9.17 (#5568) * [FrontendSpec] Adding TTLCache to `try_get_grafana_service_url` function (#5543) * [KFP] Fix `watch=True` for remote pipeline when using `set_workflow` (#5572) * [Alerts] Add cooldown_period to notification config in alert (#5547) * [Model] Replace yaml dump with safe_dump (#5571) * [Versioning] Fix package version for dev feature branches (#5575) * repace hasattr() with dir() * Support single Parquet and CSV files --------- Co-authored-by: moranbental <107995850+moranbental@users.noreply.github.com> Co-authored-by: jillnogold <88145832+jillnogold@users.noreply.github.com> Co-authored-by: Gal Topper <gal.topper@gmail.com> Co-authored-by: Yan Burman <yanburman@users.noreply.github.com> Co-authored-by: Adam <adamm@iguazio.com> Co-authored-by: Liran BG <liranbg@users.noreply.github.com> Co-authored-by: Eyal Danieli <eyald@iguazio.com> Co-authored-by: Alon Maor <48641682+alonmr@users.noreply.github.com> Co-authored-by: roei3000b <40743125+roei3000b@users.noreply.github.com> Co-authored-by: Katerina Molchanova <35141662+rokatyy@users.noreply.github.com> Co-authored-by: Jonathan Daniel <36337649+jond01@users.noreply.github.com> Co-authored-by: quaark <a.melnick@icloud.com> Co-authored-by: TomerShor <90552140+TomerShor@users.noreply.github.com> Co-authored-by: daniels290813 <78727943+daniels290813@users.noreply.github.com> Co-authored-by: ZeevRispler <73653682+ZeevRispler@users.noreply.github.com> Co-authored-by: Alex Toker <alex_toker@mckinsey.com>
Add Time Series DB abstraction that will be used for storing the model monitoring real time data. The PR includes:
tsdb
package undermlrun/model_monitoring/db/tsdb
. The abstract class calledTSDBConnector
and it includes the main applications that are being used throughout the model monitoring processes.V3IOTSDBConnector
that is stored undermlrun/model_monitoring/db/tsdb/v3io_connector
. Note that this object is based on V3IOframes
client that handles the communications with V3IO TSDB.tsdb_connector_type
. At the moment, the default value isv3io-tsdb
.Related JIRAs - https://iguazio.atlassian.net/browse/ML-5795, https://iguazio.atlassian.net/browse/ML-5796