sentry/src/sentry/api/endpoints/organization_activity.py

from functools import reduce

from rest_framework.request import Request
from rest_framework.response import Response

from sentry.api.base import EnvironmentMixin
from sentry.api.bases import OrganizationMemberEndpoint
from sentry.api.paginator import DateTimePaginator
from sentry.api.serializers import OrganizationActivitySerializer, serialize
from sentry.models import Activity, OrganizationMemberTeam, Project
from sentry.types.activity import ActivityType


class OrganizationActivityEndpoint(OrganizationMemberEndpoint, EnvironmentMixin):
    def get(self, request: Request, organization, member) -> Response:
        # There is an activity record created for both sides of the unmerge
        # operation, so we only need to include one of them here to avoid
        # showing the same entry twice.
        base_qs = Activity.objects.exclude(type=ActivityType.UNMERGE_SOURCE.value).values_list(
            "id", flat=True
        )

        # To make this query efficient, we have to hammer it into a weird format. This table is
        # extremely large and if we are making a query across many projects, in a lot of cases
        # Postgres decides that the best query plan is to iterate backwards on the datetime index.
        # This means it does something close to a table scan to get the results it wants - this can
        # be the case even for orgs with less than a page of activity rows, and often results in
        # queries that take > 30s, which get killed by stomper.
        # To convince Postgres to use the index on `(project_id, datetime)`, it basically needs to
        # see queries that filter on a single project id and then order by datetime. So we replicate
        # the query for every project and UNION ALL them together to get the candidate set of rows.
        # Then we sort these and return the final result. Convoluted, but it improves the query a
        # lot.
        # To make this work well with pagination, we have to also apply the pagination queries to
        # the subqueries.
        cursor = self.get_cursor_from_request(request)
        paginator = DateTimePaginator(base_qs, order_by="-datetime")
        if cursor is not None and cursor.value:
            cursor_value = paginator.value_from_cursor(cursor)
        else:
            cursor_value = 0
        base_qs = paginator.build_queryset(cursor_value, False)

        project_ids = list(
            Project.objects.filter(
                organization=organization,
                teams__in=OrganizationMemberTeam.objects.filter(organizationmember=member).values(
                    "team"
                ),
            ).values_list("id", flat=True)
        )

        union_qs = Activity.objects.none()
        if project_ids:
            union_qs = reduce(
                lambda qs1, qs2: qs1.union(qs2, all=True),
                [
                    base_qs.filter(project_id=project)[: paginator.max_limit]
                    for project in project_ids
                ],
            )

        # We do `select_related` here to make the unions less heavy. This way we only join these
        # table for the rows we actually want.
        queryset = Activity.objects.filter(id__in=union_qs[: paginator.max_limit]).select_related(
            "project", "group", "user"
        )

        return self.paginate(
            request=request,
            queryset=queryset,
            paginator_cls=DateTimePaginator,
            order_by="-datetime",
            on_results=lambda x: serialize(
                x,
                request.user,
                OrganizationActivitySerializer(
                    environment_func=self._get_environment_func(request, organization.id)
                ),
            ),
        )