fix: do not use grouping sets when there is no tags for group by (#4237)

This commit is contained in:
Srikanth Chekuri 2023-12-15 18:56:05 +05:30 committed by GitHub
parent cb1a823f91
commit 8e653f9500
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 13 additions and 5 deletions

View File

@ -328,7 +328,11 @@ func buildMetricQuery(start, end, step int64, mq *v3.BuilderQuery, tableName str
// `ts` is always added to the group by clause
func groupingSets(tags ...string) string {
withTs := append(tags, "ts")
return fmt.Sprintf(`GROUPING SETS ( (%s), (%s) )`, strings.Join(withTs, ", "), strings.Join(tags, ", "))
if len(withTs) > 1 {
return fmt.Sprintf(`GROUPING SETS ( (%s), (%s) )`, strings.Join(withTs, ", "), strings.Join(tags, ", "))
} else {
return strings.Join(withTs, ", ")
}
}
// groupBy returns a string of comma separated tags for group by clause

View File

@ -245,7 +245,7 @@ func TestBuildQueryOperators(t *testing.T) {
func TestBuildQueryXRate(t *testing.T) {
t.Run("TestBuildQueryXRate", func(t *testing.T) {
tmpl := `SELECT ts, %s(rate_value) as value FROM (SELECT ts, If((value - lagInFrame(value, 1, 0) OVER rate_window) < 0, nan, If((ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window) >= 86400, nan, (value - lagInFrame(value, 1, 0) OVER rate_window) / (ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window))) as rate_value FROM(SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(timestamp_ms, 1000)), INTERVAL 60 SECOND) as ts, max(value) as value FROM signoz_metrics.distributed_samples_v2 INNER JOIN (SELECT fingerprint FROM signoz_metrics.time_series_v2 WHERE metric_name = 'name' AND temporality IN ['Cumulative', 'Unspecified']) as filtered_time_series USING fingerprint WHERE metric_name = 'name' AND timestamp_ms >= 1650991980000 AND timestamp_ms <= 1651078380000 GROUP BY fingerprint, ts ORDER BY fingerprint, ts) WINDOW rate_window as (PARTITION BY fingerprint ORDER BY fingerprint, ts) ) WHERE isNaN(rate_value) = 0 GROUP BY GROUPING SETS ( (ts), () ) ORDER BY ts`
tmpl := `SELECT ts, %s(rate_value) as value FROM (SELECT ts, If((value - lagInFrame(value, 1, 0) OVER rate_window) < 0, nan, If((ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window) >= 86400, nan, (value - lagInFrame(value, 1, 0) OVER rate_window) / (ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window))) as rate_value FROM(SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(timestamp_ms, 1000)), INTERVAL 60 SECOND) as ts, max(value) as value FROM signoz_metrics.distributed_samples_v2 INNER JOIN (SELECT fingerprint FROM signoz_metrics.time_series_v2 WHERE metric_name = 'name' AND temporality IN ['Cumulative', 'Unspecified']) as filtered_time_series USING fingerprint WHERE metric_name = 'name' AND timestamp_ms >= 1650991980000 AND timestamp_ms <= 1651078380000 GROUP BY fingerprint, ts ORDER BY fingerprint, ts) WINDOW rate_window as (PARTITION BY fingerprint ORDER BY fingerprint, ts) ) WHERE isNaN(rate_value) = 0 GROUP BY ts ORDER BY ts`
cases := []struct {
aggregateOperator v3.AggregateOperator
@ -298,7 +298,7 @@ func TestBuildQueryXRate(t *testing.T) {
func TestBuildQueryRPM(t *testing.T) {
t.Run("TestBuildQueryXRate", func(t *testing.T) {
tmpl := `SELECT ts, ceil(value * 60) as value FROM (SELECT ts, %s(rate_value) as value FROM (SELECT ts, If((value - lagInFrame(value, 1, 0) OVER rate_window) < 0, nan, If((ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window) >= 86400, nan, (value - lagInFrame(value, 1, 0) OVER rate_window) / (ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window))) as rate_value FROM(SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(timestamp_ms, 1000)), INTERVAL 60 SECOND) as ts, max(value) as value FROM signoz_metrics.distributed_samples_v2 INNER JOIN (SELECT fingerprint FROM signoz_metrics.time_series_v2 WHERE metric_name = 'name' AND temporality IN ['Cumulative', 'Unspecified']) as filtered_time_series USING fingerprint WHERE metric_name = 'name' AND timestamp_ms >= 1650991980000 AND timestamp_ms <= 1651078380000 GROUP BY fingerprint, ts ORDER BY fingerprint, ts) WINDOW rate_window as (PARTITION BY fingerprint ORDER BY fingerprint, ts) ) WHERE isNaN(rate_value) = 0 GROUP BY GROUPING SETS ( (ts), () ) ORDER BY ts)`
tmpl := `SELECT ts, ceil(value * 60) as value FROM (SELECT ts, %s(rate_value) as value FROM (SELECT ts, If((value - lagInFrame(value, 1, 0) OVER rate_window) < 0, nan, If((ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window) >= 86400, nan, (value - lagInFrame(value, 1, 0) OVER rate_window) / (ts - lagInFrame(ts, 1, toDate('1970-01-01')) OVER rate_window))) as rate_value FROM(SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(timestamp_ms, 1000)), INTERVAL 60 SECOND) as ts, max(value) as value FROM signoz_metrics.distributed_samples_v2 INNER JOIN (SELECT fingerprint FROM signoz_metrics.time_series_v2 WHERE metric_name = 'name' AND temporality IN ['Cumulative', 'Unspecified']) as filtered_time_series USING fingerprint WHERE metric_name = 'name' AND timestamp_ms >= 1650991980000 AND timestamp_ms <= 1651078380000 GROUP BY fingerprint, ts ORDER BY fingerprint, ts) WINDOW rate_window as (PARTITION BY fingerprint ORDER BY fingerprint, ts) ) WHERE isNaN(rate_value) = 0 GROUP BY ts ORDER BY ts)`
cases := []struct {
aggregateOperator v3.AggregateOperator

View File

@ -11,7 +11,11 @@ import (
// `ts` is always added to the group by clause
func groupingSets(tags ...string) string {
withTs := append(tags, "ts")
return fmt.Sprintf(`GROUPING SETS ( (%s), (%s) )`, strings.Join(withTs, ", "), strings.Join(tags, ", "))
if len(withTs) > 1 {
return fmt.Sprintf(`GROUPING SETS ( (%s), (%s) )`, strings.Join(withTs, ", "), strings.Join(tags, ", "))
} else {
return strings.Join(withTs, ", ")
}
}
// groupingSetsByAttributeKeyTags returns a string of comma separated tags for group by clause

View File

@ -168,7 +168,7 @@ func TestPrepareTimeseriesQuery(t *testing.T) {
},
start: 1701794980000,
end: 1701796780000,
expectedQueryContains: "SELECT ts, sum(per_series_value) as value FROM (SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(timestamp_ms, 1000)), INTERVAL 60 SECOND) as ts, avg(value) as per_series_value FROM signoz_metrics.distributed_samples_v2 INNER JOIN (SELECT DISTINCT fingerprint FROM signoz_metrics.time_series_v2 WHERE metric_name = 'system_memory_usage' AND temporality = 'Unspecified' AND JSONExtractString(labels, 'state') != 'idle') as filtered_time_series USING fingerprint WHERE metric_name = 'system_memory_usage' AND timestamp_ms >= 1701794980000 AND timestamp_ms <= 1701796780000 GROUP BY fingerprint, ts ORDER BY fingerprint, ts) WHERE isNaN(per_series_value) = 0 GROUP BY GROUPING SETS ( (ts), () ) ORDER BY ts ASC",
expectedQueryContains: "SELECT ts, sum(per_series_value) as value FROM (SELECT fingerprint, toStartOfInterval(toDateTime(intDiv(timestamp_ms, 1000)), INTERVAL 60 SECOND) as ts, avg(value) as per_series_value FROM signoz_metrics.distributed_samples_v2 INNER JOIN (SELECT DISTINCT fingerprint FROM signoz_metrics.time_series_v2 WHERE metric_name = 'system_memory_usage' AND temporality = 'Unspecified' AND JSONExtractString(labels, 'state') != 'idle') as filtered_time_series USING fingerprint WHERE metric_name = 'system_memory_usage' AND timestamp_ms >= 1701794980000 AND timestamp_ms <= 1701796780000 GROUP BY fingerprint, ts ORDER BY fingerprint, ts) WHERE isNaN(per_series_value) = 0 GROUP BY ts ORDER BY ts ASC",
},
{
name: "test time aggregation = rate, space aggregation = sum, temporality = cumulative",