Column("timestamp", DateTime()),
Column("tags", Nested([Column("key", UInt(64)),
Column("value", UInt(64))])),
]
POST_VALUES_BUCKETS_COLUMNS: Sequence[Column[Modifiers]] = [
Column("materialization_version", UInt(8)),
Column("retention_days", UInt(16)),
Column("partition", UInt(16)),
Column("offset", UInt(64)),
]
COL_SCHEMA_DISTRIBUTIONS: Sequence[Column[Modifiers]] = [
Column(
"percentiles",
AggregateFunction("quantiles(0.5, 0.75, 0.9, 0.95, 0.99)",
[Float(64)]),
),
Column("min", AggregateFunction("min", [Float(64)])),
Column("max", AggregateFunction("max", [Float(64)])),
Column("avg", AggregateFunction("avg", [Float(64)])),
Column("sum", AggregateFunction("sum", [Float(64)])),
Column("count", AggregateFunction("count", [Float(64)])),
]
COL_SCHEMA_DISTRIBUTIONS_V2: Sequence[Column[Modifiers]] = [
*COL_SCHEMA_DISTRIBUTIONS,
Column("histogram", AggregateFunction("histogram(250)", [Float(64)])),
]
def get_forward_bucket_table_local(
raw_schema = WritableTableSchema(
columns=all_columns,
local_table_name=WRITE_LOCAL_TABLE_NAME,
dist_table_name=WRITE_DIST_TABLE_NAME,
storage_set_key=StorageSetKey.SESSIONS,
)
read_columns = ColumnSet([
("org_id", UInt(64)),
("project_id", UInt(64)),
("started", DateTime()),
("release", String()),
("environment", String()),
(
"duration_quantiles",
AggregateFunction("quantilesIf(0.5, 0.9)", UInt(32), UInt(8)),
),
("sessions", AggregateFunction("countIf", UUID(), UInt(8))),
("users", AggregateFunction("uniqIf", UUID(), UInt(8))),
(
"sessions_crashed",
AggregateFunction("countIf", UUID(), UInt(8)),
),
(
"sessions_abnormal",
AggregateFunction("countIf", UUID(), UInt(8)),
),
("sessions_errored", AggregateFunction("uniqIf", UUID(), UInt(8))),
("users_crashed", AggregateFunction("uniqIf", UUID(), UInt(8))),
("users_abnormal", AggregateFunction("uniqIf", UUID(), UInt(8))),
("users_errored", AggregateFunction("uniqIf", UUID(), UInt(8))),
test_data = [
# Basic types
(("Date", "", "", ""), Date()),
(("DateTime", "", "", ""), DateTime()),
(("Enum8('success' = 0, 'error' = 1)", "", "", ""),
Enum([("success", 0), ("error", 1)])),
(("FixedString(32)", "", "", ""), FixedString(32)),
(("Float32", "", "", ""), Float(32)),
(("IPv4", "", "", ""), IPv4()),
(("IPv6", "", "", ""), IPv6()),
(("String", "", "", ""), String()),
(("UInt32", "", "", ""), UInt(32)),
(("UUID", "", "", ""), UUID()),
# Aggregate functions
(("AggregateFunction(uniq, UInt8)", "", "", ""),
AggregateFunction("uniq", UInt(8))),
(("AggregateFunction(countIf, UUID, UInt8)", "", "", ""),
AggregateFunction("countIf", UUID(), UInt(8))),
(("AggregateFunction(quantileIf(0.5, 0.9), UInt32, UInt8)", "", "", ""),
AggregateFunction("quantileIf(0.5, 0.9)", UInt(32), UInt(8))),
# Array
(("Array(String)", "", "", ""), Array(String())),
(("Array(DateTime)", "", "", ""), Array(DateTime())),
(("Array(UInt64)", "", "", ""), Array(UInt(64))),
(("Array(Nullable(UUID))", "", "", ""), Array(Nullable(UUID()))),
# Nullable
(("Nullable(String)", "", "", ""), Nullable(String())),
(("Nullable(FixedString(8))", "", "", ""), Nullable(FixedString(8))),
(("Nullable(Date)", "", "", ""), Nullable(Date())),
# Low cardinality
(("LowCardinality(String)", "", "", ""), LowCardinality(String())),
class Migration(migration.ClickhouseNodeMigration):
blocking = False
view_name = "generic_metric_sets_aggregation_mv"
dest_table_columns: Sequence[Column[Modifiers]] = [
Column("org_id", UInt(64)),
Column("project_id", UInt(64)),
Column("metric_id", UInt(64)),
Column("granularity", UInt(8)),
Column("timestamp",
DateTime(modifiers=Modifiers(codecs=["DoubleDelta"]))),
Column("retention_days", UInt(16)),
Column(
"tags",
Nested([
("key", UInt(64)),
("indexed_value", UInt(64)),
("raw_value", String()),
]),
),
Column("value", AggregateFunction("uniqCombined64", [UInt(64)])),
Column("use_case_id", String(Modifiers(low_cardinality=True))),
]
def forwards_local(self) -> Sequence[operations.SqlOperation]:
return [
operations.DropTable(
storage_set=StorageSetKey.GENERIC_METRICS_SETS,
table_name=self.view_name,
),
operations.CreateMaterializedView(
storage_set=StorageSetKey.GENERIC_METRICS_SETS,
view_name=self.view_name,
columns=self.dest_table_columns,
destination_table_name="generic_metric_sets_local",
query="""
SELECT
use_case_id,
org_id,
project_id,
metric_id,
arrayJoin(granularities) as granularity,
tags.key,
tags.indexed_value,
tags.raw_value,
toDateTime(multiIf(granularity=0,10,granularity=1,60,granularity=2,3600,granularity=3,86400,-1) *
intDiv(toUnixTimestamp(timestamp),
multiIf(granularity=0,10,granularity=1,60,granularity=2,3600,granularity=3,86400,-1))) as timestamp,
retention_days,
uniqCombined64State(arrayJoin(set_values)) as value
FROM generic_metric_sets_raw_local
WHERE materialization_version = 1
AND metric_type = 'set'
GROUP BY
use_case_id,
org_id,
project_id,
metric_id,
tags.key,
tags.indexed_value,
tags.raw_value,
timestamp,
granularity,
retention_days
""",
),
]
def backwards_local(self) -> Sequence[operations.SqlOperation]:
return [
operations.DropTable(
storage_set=StorageSetKey.GENERIC_METRICS_SETS,
table_name=self.view_name,
)
]
def forwards_dist(self) -> Sequence[operations.SqlOperation]:
return []
def backwards_dist(self) -> Sequence[operations.SqlOperation]:
return []
from snuba.migrations.columns import MigrationModifiers as Modifiers
from snuba.processor import MAX_UINT32, NIL_UUID
from .matview import create_matview_v1
aggregate_columns: Sequence[Column[Modifiers]] = [
Column("org_id", UInt(64)),
Column("project_id", UInt(64)),
Column("started", DateTime()),
Column("release", String(Modifiers(low_cardinality=True))),
Column("environment", String(Modifiers(low_cardinality=True))),
Column("user_agent", String(Modifiers(low_cardinality=True))),
Column("os", String(Modifiers(low_cardinality=True))),
# durations
Column(
"duration_quantiles",
AggregateFunction("quantilesIf(0.5, 0.9)",
[UInt(32), UInt(8)]),
),
Column("duration_avg", AggregateFunction("avgIf",
[UInt(32), UInt(8)])),
# sessions:
Column("sessions", AggregateFunction("countIf", [UUID(), UInt(8)])),
Column("sessions_preaggr", AggregateFunction("sumIf",
[UInt(32), UInt(8)])),
Column("sessions_crashed", AggregateFunction("countIf",
[UUID(), UInt(8)])),
Column("sessions_crashed_preaggr",
AggregateFunction("sumIf", [UInt(32), UInt(8)])),
Column("sessions_abnormal", AggregateFunction("countIf",
[UUID(), UInt(8)])),
Column("sessions_abnormal_preaggr",
AggregateFunction("sumIf", [UInt(32), UInt(8)])),
[("key", String()), ("val", String(Modifier(nullable=True)))],
Modifier(nullable=True),
),
Nested([("key", String()), ("val", String())]),
cast(
Column[Modifier],
Nested([("key", String()), ("val", String())],
Modifier(nullable=True)),
),
"Nullable(Nested(key String, val Nullable(String)))",
id="nested",
),
pytest.param(
cast(
Column[Modifier],
AggregateFunction("uniqIf", [UInt(8), UInt(32)],
Modifier(nullable=True)),
),
AggregateFunction("uniqIf", [UInt(8), UInt(32)]),
cast(
Column[Modifier],
AggregateFunction("uniqIf", [UInt(8)], Modifier(nullable=True)),
),
"Nullable(AggregateFunction(uniqIf, UInt8, UInt32))",
id="aggregated",
),
pytest.param(
Enum([("a", 1), ("b", 2)], Modifier(nullable=True)),
Enum([("a", 1), ("b", 2)]),
Enum([("a", 1), ("b", 2)]),
"Nullable(Enum('a' = 1, 'b' = 2))",
id="enums",
new_raw_columns: Sequence[Tuple[Column[Modifiers], str]] = [
(
Column("quantity", UInt(32, Modifiers(default=str(DEFAULT_QUANTITY)))),
"distinct_id",
),
(Column("user_agent",
String(Modifiers(low_cardinality=True))), "environment"),
(Column("os", String(Modifiers(low_cardinality=True))), "user_agent"),
]
new_dest_columns: Sequence[Tuple[Column[Modifiers], str]] = [
(Column("user_agent",
String(Modifiers(low_cardinality=True))), "environment"),
(Column("os", String(Modifiers(low_cardinality=True))), "user_agent"),
(
Column("duration_avg", AggregateFunction("avgIf",
[UInt(32), UInt(8)])),
"duration_quantiles",
),
(
Column("sessions_preaggr",
AggregateFunction("sumIf", [UInt(32), UInt(8)])),
"sessions",
),
(
Column(
"sessions_crashed_preaggr",
AggregateFunction("sumIf", [UInt(32), UInt(8)]),
),
"sessions_crashed",
),
(
Column("timestamp", DateTime()),
Column("retention_days", UInt(16)),
Column("tags", Nested([("key", UInt(64)), ("value", UInt(64))])),
Column("_tags_hash", Array(UInt(64), SchemaModifiers(readonly=True))),
]
sets_storage = ReadableTableStorage(
storage_key=StorageKey.METRICS_SETS,
storage_set_key=StorageSetKey.METRICS,
schema=TableSchema(
local_table_name="metrics_sets_local",
dist_table_name="metrics_sets_dist",
storage_set_key=StorageSetKey.METRICS,
columns=ColumnSet([
*aggregated_columns,
Column("value", AggregateFunction("uniqCombined64", [UInt(64)])),
]),
),
query_processors=[ArrayJoinKeyValueOptimizer("tags")],
)
counters_storage = ReadableTableStorage(
storage_key=StorageKey.METRICS_COUNTERS,
storage_set_key=StorageSetKey.METRICS,
schema=TableSchema(
local_table_name="metrics_counters_local",
dist_table_name="metrics_counters_dist",
storage_set_key=StorageSetKey.METRICS,
columns=ColumnSet([
*aggregated_columns,
Column("value", AggregateFunction("sum", [Float(64)])),
def visit_agg(self, node: Node,
visited_children: Iterable[Any]) -> AggregateFunction:
(_agg, _paren, _sp, agg_func, _sp, _comma, _sp, agg_types, _sp,
_paren) = visited_children
return AggregateFunction(agg_func, *agg_types)
