feat: add auto update script for all datasets in marked channels #178 (#181)

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Author: Fedir-Yatsenko

Makefile

@@ -45,6 +45,12 @@ statgpt_cli: install_dev
 statgpt_admin:
 	poetry run python -m statgpt.admin.app $(ARGS)
 
+statgpt_fix_statuses:
+	poetry run python -m statgpt.admin.fix_statuses
+
+statgpt_auto_update:
+	poetry run python -m statgpt.admin.auto_update
+
 statgpt_app:
 	poetry run python -m statgpt.app.app $(ARGS)
 

statgpt/admin/admin.sh

@@ -26,12 +26,17 @@ case "${ADMIN_MODE:-}" in
     python -m statgpt.admin.fix_statuses
     ;;
 
+  AUTO_UPDATE)
+    python -m statgpt.admin.auto_update
+    ;;
+
   *)
     echo "Unknown ADMIN_MODE = '${ADMIN_MODE:-}'. Possible values:"
     echo "  APP - start the admin application"
     echo "  ALEMBIC_UPGRADE - run alembic migrations to upgrade the database"
     echo "  FIX_STATUSES - fix inconsistent statuses in the database"
     echo "  INIT - run alembic migrations and fix inconsistent statuses"
+    echo "  AUTO_UPDATE - run batch auto-update for all eligible channels"
     exit 1
     ;;
 esac

statgpt/admin/auto_update.py

@@ -0,0 +1,154 @@
+"""
+Batch auto-update script for all datasets in channels with `allow_auto_update` enabled.
+"""
+
+import asyncio
+import logging
+import sys
+
+import statgpt.common.schemas as schemas
+from statgpt.admin.auth.auth_context import SystemUserAuthContext
+from statgpt.admin.services.channel import (
+    AdminPortalChannelService,
+    deduplicate_dimensions_in_background_task,
+)
+from statgpt.admin.services.dataset import AdminPortalDataSetService, auto_update_in_background_task
+from statgpt.common.auth.auth_context import AuthContext
+from statgpt.common.models import get_session_contex_manager, optional_msi_token_manager_context
+
+_log = logging.getLogger(__name__)
+_SEPARATOR = "-" * 50
+
+
+async def _discover_and_create_jobs() -> list[schemas.AutoUpdateJob]:
+    """Find auto-update channels and create jobs for their datasets."""
+    _log.info(_SEPARATOR)
+    async with get_session_contex_manager() as session:
+        channel_service = AdminPortalChannelService(session)
+        all_channels = await channel_service.get_channels_schemas(limit=None, offset=0)
+        channel_ids = [
+            ch.id
+            for ch in all_channels
+            if (dq := ch.details.data_query) is not None and dq.details.allow_auto_update
+        ]
+        _log.info(f"Found {len(channel_ids)} channel(s) with auto-update enabled")
+
+        if not channel_ids:
+            return []
+
+        return await AdminPortalDataSetService(session).create_auto_update_jobs(channel_ids)
+
+
+async def _process_jobs(jobs: list[schemas.AutoUpdateJob], auth_context: AuthContext) -> None:
+    """Run all auto-update jobs concurrently.
+
+    NOTE: The number of concurrent executions is limited by the semaphore
+    in the ``@background_task`` decorator applied to ``auto_update_in_background_task``.
+    """
+    _log.info(_SEPARATOR)
+    _log.info(f"Created {len(jobs)} auto-update job(s), starting processing...")
+
+    results = await asyncio.gather(
+        *(
+            auto_update_in_background_task(auto_update_job_id=job.id, auth_context=auth_context)
+            for job in jobs
+        ),
+        return_exceptions=True,
+    )
+    for job, result in zip(jobs, results):
+        if isinstance(result, Exception):
+            _log.error(f"Auto-update job {job.id} failed with exception:", exc_info=result)
+
+
+async def _get_reindex_channel_ids(job_ids: list[int]) -> set[int]:
+    """Get channel IDs that had at least one reindex triggered."""
+    async with get_session_contex_manager() as session:
+        return await AdminPortalDataSetService(session).get_reindex_channel_ids(job_ids)
+
+
+async def _log_results(job_ids: list[int]) -> bool:
+    """Log per-channel summary and return `True` if all jobs succeeded."""
+    _log.info(_SEPARATOR)
+    async with get_session_contex_manager() as session:
+        results = await AdminPortalDataSetService(session).get_auto_update_results(job_ids)
+
+    for r in results:
+        _log.info(f"channel '{r.deployment_id}' (id={r.channel_id}): {r.summary}")
+        for reason in r.failed_reasons:
+            _log.error(f"  {reason}")
+
+    total = sum(r.total for r in results)
+    failed = sum(r.failed for r in results)
+    _log.info(
+        f"Auto-update complete: {total - failed} succeeded, {failed} failed "
+        f"out of {total} total"
+    )
+    return failed == 0
+
+
+async def _deduplicate_channels(channel_ids: set[int], auth_context: AuthContext) -> None:
+    """Run deduplication for channels that had a reindex.
+
+    NOTE: The number of concurrent executions is limited by the semaphore
+    in the ``@background_task`` decorator applied to ``deduplicate_dimensions_in_background_task``.
+    """
+    _log.info(_SEPARATOR)
+    sorted_ids = sorted(channel_ids)
+    _log.info(f"Running deduplication for {len(sorted_ids)} channel(s) with reindex: {sorted_ids}")
+    results = await asyncio.gather(
+        *(
+            deduplicate_dimensions_in_background_task(
+                channel_id=channel_id, auth_context=auth_context
+            )
+            for channel_id in sorted_ids
+        ),
+        return_exceptions=True,
+    )
+    for channel_id, result in zip(sorted_ids, results):
+        if isinstance(result, Exception):
+            _log.error(
+                f"Deduplication for channel {channel_id} failed with exception:", exc_info=result
+            )
+    _log.info("Deduplication complete")
+
+
+async def run_auto_update() -> bool:
+    """Run batch auto-update for all eligible channels.
+
+    Returns:
+         `True` if all jobs succeeded, `False` otherwise.
+    """
+    auth_context = SystemUserAuthContext()
+
+    jobs = await _discover_and_create_jobs()
+    if not jobs:
+        return True
+
+    await _process_jobs(jobs, auth_context)
+    job_ids = [j.id for j in jobs]
+
+    reindex_channel_ids = await _get_reindex_channel_ids(job_ids)
+    if reindex_channel_ids:
+        await _deduplicate_channels(reindex_channel_ids, auth_context)
+
+    return await _log_results(job_ids)
+
+
+async def main() -> None:
+    try:
+        _log.info("Starting batch auto-update script...")
+        async with optional_msi_token_manager_context():
+            success = await run_auto_update()
+
+        _log.info(_SEPARATOR)
+        if not success:
+            _log.error("Batch auto-update finished with failures")
+            sys.exit(1)
+        _log.info("Batch auto-update script completed successfully")
+    except Exception:
+        _log.exception("Error in batch auto-update script:")
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

statgpt/admin/services/dataset.py

@@ -3,13 +3,15 @@
 import os.path
 import uuid
 import zipfile
+from collections import Counter, defaultdict
 from collections.abc import Generator, Iterable
 from typing import Any, NamedTuple
 
 import yaml
 from fastapi import BackgroundTasks, HTTPException, status
 from pydantic import ValidationError
 from sqlalchemy.ext.asyncio import AsyncSession
+from sqlalchemy.orm import selectinload
 from sqlalchemy.sql.expression import func, select, text, update
 
 import statgpt.common.models as models
@@ -25,6 +27,7 @@
 from statgpt.common.schemas import (
     AuditActionType,
     AuditEntityType,
+    AutoUpdateResult,
     ChannelIndexStatusScope,
     HybridSearchConfig,
 )
@@ -59,6 +62,15 @@ class _DataHashes(NamedTuple):
     special_dimensions_hash: str | None
 
 
+class AutoUpdateChannelResult(NamedTuple):
+    channel_id: int
+    deployment_id: str
+    total: int
+    failed: int
+    summary: str
+    failed_reasons: list[str]
+
+
 class AdminPortalDataSetService(DataSetService):
 
     def __init__(self, session: AsyncSession) -> None:
@@ -1989,6 +2001,105 @@ async def get_auto_update_jobs_count(self, channel_dataset_id: int) -> int:
         )
         return await self._session.scalar(query) or 0
 
+    async def create_auto_update_jobs(self, channel_ids: list[int]) -> list[schemas.AutoUpdateJob]:
+        """Create AutoUpdateJob records for all datasets in the given channels.
+
+        Returns a list of created job schemas.
+        """
+        result = await self._session.execute(
+            select(models.ChannelDataset)
+            .where(models.ChannelDataset.channel_id.in_(channel_ids))
+            .options(selectinload(models.ChannelDataset.channel))
+        )
+        channel_datasets = list(result.scalars().all())
+
+        jobs: list[models.AutoUpdateJob] = []
+        for cd in channel_datasets:
+            job = models.AutoUpdateJob(channel_dataset_id=cd.id, status=StatusEnum.QUEUED)
+            self._session.add(job)
+            jobs.append(job)
+        await self._session.commit()
+
+        # Log per-channel summary
+        channels_by_id = {cd.channel.id: cd.channel for cd in channel_datasets}
+        counts = Counter(cd.channel_id for cd in channel_datasets)
+        for ch_id, count in counts.items():
+            ch = channels_by_id[ch_id]
+            _log.info(
+                f"Created {count} auto-update job(s) for channel '{ch.deployment_id}' (id={ch_id})"
+            )
+
+        return [schemas.AutoUpdateJob.model_validate(job, from_attributes=True) for job in jobs]
+
+    async def get_reindex_channel_ids(self, job_ids: list[int]) -> set[int]:
+        """Get channel IDs that had at least one REINDEX_TRIGGERED result."""
+        result = await self._session.execute(
+            select(models.AutoUpdateJob)
+            .where(models.AutoUpdateJob.id.in_(job_ids))
+            .where(models.AutoUpdateJob.result == AutoUpdateResult.REINDEX_TRIGGERED)
+            .options(selectinload(models.AutoUpdateJob.channel_dataset))
+        )
+        jobs = list(result.scalars().all())
+        return {j.channel_dataset.channel_id for j in jobs}
+
+    async def get_auto_update_results(self, job_ids: list[int]) -> list[AutoUpdateChannelResult]:
+        """Collect per-channel auto-update results."""
+        result = await self._session.execute(
+            select(models.AutoUpdateJob)
+            .where(models.AutoUpdateJob.id.in_(job_ids))
+            .options(
+                selectinload(models.AutoUpdateJob.channel_dataset).selectinload(
+                    models.ChannelDataset.channel
+                ),
+                selectinload(models.AutoUpdateJob.created_version),
+            )
+        )
+        jobs = list(result.scalars().all())
+
+        channel_jobs: defaultdict[int, list[models.AutoUpdateJob]] = defaultdict(list)
+        for job in jobs:
+            channel_jobs[job.channel_dataset.channel.id].append(job)
+
+        results: list[AutoUpdateChannelResult] = []
+        for channel_id, ch_jobs in channel_jobs.items():
+            ch = ch_jobs[0].channel_dataset.channel
+            results.append(
+                AutoUpdateChannelResult(
+                    channel_id=channel_id,
+                    deployment_id=ch.deployment_id,
+                    total=len(ch_jobs),
+                    failed=sum(1 for j in ch_jobs if j.status == StatusEnum.FAILED),
+                    summary=self._format_result_summary(ch_jobs),
+                    failed_reasons=[
+                        f"job {j.id}: {j.reason_for_failure}"
+                        for j in ch_jobs
+                        if j.status == StatusEnum.FAILED
+                    ],
+                )
+            )
+        return results
+
+    @staticmethod
+    def _format_result_summary(jobs: list[models.AutoUpdateJob]) -> str:
+        """Build a human-readable summary of auto-update results."""
+        reindex_statuses: dict[str, Counter[str]] = defaultdict(Counter)
+        job_statuses: list[str] = []
+        for job in jobs:
+            job_status = job.result.value if job.result else job.status.value
+            job_statuses.append(job_status)
+            if job.created_version is not None:
+                reindex_statuses[job_status][job.created_version.preprocessing_status.value] += 1
+        result_counts = Counter(job_statuses)
+
+        parts: list[str] = []
+        for job_status, count in result_counts.most_common():
+            part = f"{count} {job_status}"
+            if job_status in reindex_statuses:
+                breakdown = ", ".join(f"{c} {s}" for s, c in reindex_statuses[job_status].items())
+                part += f" ({breakdown})"
+            parts.append(part)
+        return ", ".join(parts)
+
     async def _set_auto_update_job_status(
         self,
         job: models.AutoUpdateJob,

statgpt/common/schemas/data_query_tool.py

@@ -268,6 +268,10 @@ class DataQueryDetails(BaseToolDetails):
     prompts: DataQueryPrompts = Field(default_factory=DataQueryPrompts)  # type: ignore
     messages: DataQueryMessages = Field(default_factory=DataQueryMessages)  # type: ignore
     attachments: DataQueryAttachments = Field(default_factory=DataQueryAttachments)  # type: ignore
+    allow_auto_update: bool = Field(
+        default=False,
+        description="Whether datasets in this channel should be auto-updated by the batch auto-update script.",
+    )
     tool_response_max_cells: PositiveInt = Field(
         default=300,
         description=(

statgpt/common/services/channel.py

@@ -27,7 +27,7 @@ async def get_channels_db(self, limit: int | None, offset: int) -> list[models.C
         q_result = await self._session.execute(query)
         return [item for item in q_result.scalars().all()]
 
-    async def get_channels_schemas(self, limit: int, offset: int) -> list[schemas.Channel]:
+    async def get_channels_schemas(self, limit: int | None, offset: int) -> list[schemas.Channel]:
         channels = await self.get_channels_db(limit, offset)
         return [ChannelSerializer.db_to_schema(item) for item in channels]