Skip to content

perf: add prelimit CTE to getMapKeys query + store clickhouse settings in shared cache #1187

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.

Sign up for GitHub

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

Closed
wrn14897 wants to merge 4 commits into from
Closed

perf: add prelimit CTE to getMapKeys query + store clickhouse settings in shared cache #1187

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
13d7ff8
perf: add prelimit CTE to getMapKeys query (#611)
wrn14897 Aug 8, 2025
435bb6e
fix + perf: store clickhouse settings in shared cache (#672)
wrn14897 Sep 4, 2025
dad982a
docs: add changeset
wrn14897 Sep 19, 2025
5c7a5e2
Merge branch 'main' into warren/opt-metadata-field-queries
wrn14897 Sep 23, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions .changeset/strange-boats-nail.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
---
"@hyperdx/common-utils": patch
"@hyperdx/app": patch
---

perf: add prelimit CTE to getMapKeys query + store clickhouse settings in shared cache
4 changes: 2 additions & 2 deletions packages/app/src/components/DBSearchPageFilters.tsx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,8 +33,8 @@ import {
useJsonColumns,
useTableMetadata,
} from '@/hooks/useMetadata';
import { useMetadataWithSettings } from '@/hooks/useMetadata';
import useResizable from '@/hooks/useResizable';
import { getMetadata } from '@/metadata';
import { FilterStateHook, usePinnedFilters } from '@/searchFilters';
import { useSource } from '@/source';
import { mergePath } from '@/utils';
Expand Down Expand Up @@ -615,6 +615,7 @@ const DBSearchPageFiltersComponent = ({
keys: keysToFetch,
});

const metadata = useMetadataWithSettings();
const [extraFacets, setExtraFacets] = useState<Record<string, string[]>>({});
const [loadMoreLoadingKeys, setLoadMoreLoadingKeys] = useState<Set<string>>(
new Set(),
Expand All @@ -623,7 +624,6 @@ const DBSearchPageFiltersComponent = ({
async (key: string) => {
setLoadMoreLoadingKeys(prev => new Set(prev).add(key));
try {
const metadata = getMetadata();
const newKeyVals = await metadata.getKeyValues({
chartConfig: {
...chartConfig,
Expand Down
48 changes: 26 additions & 22 deletions packages/app/src/hooks/useMetadata.tsx
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
import { useEffect, useRef } from 'react';
import objectHash from 'object-hash';
import {
ColumnMeta,
Expand All @@ -20,6 +21,25 @@ import api from '@/api';
import { getMetadata } from '@/metadata';
import { toArray } from '@/utils';

// Hook to get metadata with proper settings applied
// TODO: replace all getMetadata calls with useMetadataWithSettings
export function useMetadataWithSettings() {
const metadata = getMetadata();
const { data: me } = api.useMe();
const settingsApplied = useRef(false);

useEffect(() => {
if (me?.team?.metadataMaxRowsToRead && !settingsApplied.current) {
metadata.setClickHouseSettings({
max_rows_to_read: me.team.metadataMaxRowsToRead,
});
settingsApplied.current = true;
}
}, [me?.team?.metadataMaxRowsToRead, metadata]);

return metadata;
}

export function useColumns(
{
databaseName,
Expand All @@ -32,10 +52,10 @@ export function useColumns(
},
options?: Partial<UseQueryOptions<ColumnMeta[]>>,
) {
const metadata = useMetadataWithSettings();
return useQuery<ColumnMeta[]>({
queryKey: ['useMetadata.useColumns', { databaseName, tableName }],
queryFn: async () => {
const metadata = getMetadata();
return metadata.getColumns({
databaseName,
tableName,
Expand All @@ -51,10 +71,10 @@ export function useJsonColumns(
{ databaseName, tableName, connectionId }: TableConnection,
options?: Partial<UseQueryOptions<string[]>>,
) {
const metadata = useMetadataWithSettings();
return useQuery<string[]>({
queryKey: ['useMetadata.useJsonColumns', { databaseName, tableName }],
queryFn: async () => {
const metadata = getMetadata();
const columns = await metadata.getColumns({
databaseName,
tableName,
Expand All @@ -78,7 +98,7 @@ export function useAllFields(
const tableConnections = Array.isArray(_tableConnections)
? _tableConnections
: [_tableConnections];
const metadata = getMetadata();
const metadata = useMetadataWithSettings();
const { data: me, isFetched } = api.useMe();
return useQuery<Field[]>({
queryKey: [
Expand All @@ -91,13 +111,6 @@ export function useAllFields(
return [];
}

// TODO: set the settings at the top level so that it doesn't have to be set for each useQuery
if (team?.metadataMaxRowsToRead) {
metadata.setClickHouseSettings({
max_rows_to_read: team.metadataMaxRowsToRead,
});
}

const fields2d = await Promise.all(
tableConnections.map(tc => metadata.getAllFields(tc)),
);
Expand Down Expand Up @@ -129,7 +142,7 @@ export function useTableMetadata(
},
options?: Omit<UseQueryOptions<any, Error>, 'queryKey'>,
) {
const metadata = getMetadata();
const metadata = useMetadataWithSettings();
return useQuery<TableMetadata>({
queryKey: ['useMetadata.useTableMetadata', { databaseName, tableName }],
queryFn: async () => {
Expand Down Expand Up @@ -159,9 +172,8 @@ export function useGetKeyValues(
},
options?: Omit<UseQueryOptions<any, Error>, 'queryKey'>,
) {
const metadata = getMetadata();
const metadata = useMetadataWithSettings();
const chartConfigsArr = toArray(chartConfigs);
const { data: me, isFetched } = api.useMe();
return useQuery<{ key: string; value: string[] }[]>({
queryKey: [
'useMetadata.useGetKeyValues',
Expand All @@ -170,14 +182,6 @@ export function useGetKeyValues(
disableRowLimit,
],
queryFn: async () => {
const team = me?.team;

// TODO: set the settings at the top level so that it doesn't have to be set for each useQuery
if (team?.metadataMaxRowsToRead) {
metadata.setClickHouseSettings({
max_rows_to_read: team.metadataMaxRowsToRead,
});
}
return (
await Promise.all(
chartConfigsArr.map(chartConfig =>
Expand All @@ -192,7 +196,7 @@ export function useGetKeyValues(
).flatMap(v => v);
},
staleTime: 1000 * 60 * 5, // Cache every 5 min
enabled: !!keys.length && isFetched,
enabled: !!keys.length,
placeholderData: keepPreviousData,
...options,
});
Expand Down
109 changes: 84 additions & 25 deletions packages/common-utils/src/metadata.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,7 +94,6 @@ export type TableMetadata = {
export class Metadata {
private readonly clickhouseClient: BaseClickhouseClient;
private readonly cache: MetadataCache;
private readonly clickhouseSettings: ClickHouseSettings;

constructor(
clickhouseClient: BaseClickhouseClient,
Expand All @@ -103,11 +102,19 @@ export class Metadata {
) {
this.clickhouseClient = clickhouseClient;
this.cache = cache;
this.clickhouseSettings = settings ?? {};
if (settings) {
this.cache.set('clickhouse-settings', settings);
}
}

getClickHouseSettings(): ClickHouseSettings {
return this.cache.get<ClickHouseSettings>('clickhouse-settings') ?? {};
}

setClickHouseSettings(settings: ClickHouseSettings) {
Object.assign(this.clickhouseSettings, settings);
const currentSettings = this.getClickHouseSettings();
const updatedSettings = { ...currentSettings, ...settings };
this.cache.set('clickhouse-settings', updatedSettings);
}

private async queryTableMetadata({
Expand All @@ -128,7 +135,7 @@ export class Metadata {
connectionId,
query: sql.sql,
query_params: sql.params,
clickhouse_settings: this.clickhouseSettings,
clickhouse_settings: this.getClickHouseSettings(),
})
.then(res => res.json<TableMetadata>());
return json.data[0];
Expand All @@ -153,7 +160,7 @@ export class Metadata {
query: sql.sql,
query_params: sql.params,
connectionId,
clickhouse_settings: this.clickhouseSettings,
clickhouse_settings: this.getClickHouseSettings(),
})
.then(res => res.json())
.then(d => d.data);
Expand Down Expand Up @@ -264,18 +271,39 @@ export class Metadata {
: '';
let sql: ChSql;
if (strategy === 'groupUniqArrayArray') {
sql = chSql`SELECT groupUniqArrayArray(${{ Int32: maxKeys }})(${{
Identifier: column,
}}) as keysArr
FROM ${tableExpr({ database: databaseName, table: tableName })} ${where}`;
sql = chSql`
WITH sampledKeys as (
SELECT ${{
Identifier: column,
}}.keys AS keys
FROM ${tableExpr({ database: databaseName, table: tableName })} ${where}
LIMIT ${{
Int32: this.getClickHouseSettings().max_rows_to_read
? Number(this.getClickHouseSettings().max_rows_to_read)
: DEFAULT_METADATA_MAX_ROWS_TO_READ,
}}
)
SELECT groupUniqArrayArray(${{ Int32: maxKeys }})(keys) as keysArr
FROM sampledKeys`;
} else {
sql = chSql`SELECT DISTINCT lowCardinalityKeys(arrayJoin(${{
Identifier: column,
}}.keys)) as key
FROM ${tableExpr({ database: databaseName, table: tableName })} ${where}
LIMIT ${{
Int32: maxKeys,
}}`;
sql = chSql`
WITH sampledKeys as (
SELECT ${{
Identifier: column,
}}.keys AS keysArr
FROM ${tableExpr({ database: databaseName, table: tableName })} ${where}
LIMIT ${{
Int32: this.getClickHouseSettings().max_rows_to_read
? Number(this.getClickHouseSettings().max_rows_to_read)
: DEFAULT_METADATA_MAX_ROWS_TO_READ,
}}
)
SELECT DISTINCT lowCardinalityKeys(arrayJoin(keysArr)) as key
FROM sampledKeys
LIMIT ${{
Int32: maxKeys,
}}
`;
}

return this.cache.getOrFetch<string[]>(cacheKey, async () => {
Expand All @@ -285,9 +313,12 @@ export class Metadata {
query_params: sql.params,
connectionId,
clickhouse_settings: {
max_rows_to_read: String(DEFAULT_METADATA_MAX_ROWS_TO_READ),
max_rows_to_read: String(
this.getClickHouseSettings().max_rows_to_read ??
DEFAULT_METADATA_MAX_ROWS_TO_READ,
),
read_overflow_mode: 'break',
...this.clickhouseSettings,
...this.getClickHouseSettings(),
},
})
.then(res => res.json<Record<string, unknown>>())
Expand Down Expand Up @@ -342,9 +373,12 @@ export class Metadata {
query_params: sql.params,
connectionId,
clickhouse_settings: {
max_rows_to_read: String(DEFAULT_METADATA_MAX_ROWS_TO_READ),
max_rows_to_read: String(
this.getClickHouseSettings().max_rows_to_read ??
DEFAULT_METADATA_MAX_ROWS_TO_READ,
),
read_overflow_mode: 'break',
...this.clickhouseSettings,
...this.getClickHouseSettings(),
},
})
.then(res => res.json<{ pathMap: Record<string, string[]> }>())
Expand Down Expand Up @@ -422,9 +456,12 @@ export class Metadata {
query_params: sql.params,
connectionId,
clickhouse_settings: {
max_rows_to_read: String(DEFAULT_METADATA_MAX_ROWS_TO_READ),
max_rows_to_read: String(
this.getClickHouseSettings().max_rows_to_read ??
DEFAULT_METADATA_MAX_ROWS_TO_READ,
),
read_overflow_mode: 'break',
...this.clickhouseSettings,
...this.getClickHouseSettings(),
},
})
.then(res => res.json<Record<string, unknown>>())
Expand Down Expand Up @@ -545,10 +582,29 @@ export class Metadata {
async () => {
const sql = await renderChartConfig(
{
...chartConfig,
with: [
{
name: 'sampledData',

@MikeShi42 MikeShi42 Sep 19, 2025

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This actually introduces the regression noted in HDX-2440. Before we merge this PR we should have a fix prepared to land on top so we can smoothly fix the regression without too much weirdness in downstream git history while not breaking the next build.

@wrn14897 wrn14897 Sep 19, 2025

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sounds good. I will take a look at it

chartConfig: {
...chartConfig,
select: keys.join(', '),
limit: {
limit: !disableRowLimit
? this.getClickHouseSettings().max_rows_to_read
? Number(this.getClickHouseSettings().max_rows_to_read)
: DEFAULT_METADATA_MAX_ROWS_TO_READ
: undefined,
},
},
isSubquery: true,
},
],
select: keys
.map((k, i) => `groupUniqArray(${limit})(${k}) AS param${i}`)
.join(', '),
connection: chartConfig.connection,
from: { databaseName: '', tableName: 'sampledData' },
where: '',
},
this,
);
Expand All @@ -560,9 +616,12 @@ export class Metadata {
connectionId: chartConfig.connection,
clickhouse_settings: !disableRowLimit
? {
max_rows_to_read: String(DEFAULT_METADATA_MAX_ROWS_TO_READ),
max_rows_to_read: String(
this.getClickHouseSettings().max_rows_to_read ??
DEFAULT_METADATA_MAX_ROWS_TO_READ,
),
read_overflow_mode: 'break',
...this.clickhouseSettings,
...this.getClickHouseSettings(),
}
: undefined,
})
Expand Down