Félix Malfait
30 lines
1.1 KiB
SQL
30 lines
1.1 KiB
SQL
-- Optimize objectEvent table for time-series queries and object filtering
|
|
-- ClickHouse doesn't allow changing ORDER BY, so we create a new table and migrate data
|
|
|
|
-- Step 1: Create new table with optimized structure
|
|
CREATE TABLE IF NOT EXISTS objectEvent_v2
|
|
(
|
|
`event` LowCardinality(String) NOT NULL,
|
|
`timestamp` DateTime64(3) NOT NULL,
|
|
`userWorkspaceId` String DEFAULT '',
|
|
`workspaceId` String NOT NULL,
|
|
`recordId` String NOT NULL,
|
|
`objectMetadataId` String NOT NULL,
|
|
`properties` JSON,
|
|
`isCustom` Boolean DEFAULT FALSE
|
|
)
|
|
ENGINE = MergeTree
|
|
ORDER BY (workspaceId, timestamp, objectMetadataId, recordId, event, userWorkspaceId)
|
|
TTL timestamp + INTERVAL 3 YEAR DELETE;
|
|
|
|
-- Step 2: Migrate existing data (userWorkspaceId will be empty for historical data)
|
|
INSERT INTO objectEvent_v2
|
|
SELECT event, timestamp, '' as userWorkspaceId, workspaceId, recordId, objectMetadataId, properties, isCustom
|
|
FROM objectEvent;
|
|
|
|
-- Step 3: Atomic swap
|
|
EXCHANGE TABLES objectEvent AND objectEvent_v2;
|
|
|
|
-- Step 4: Drop old table
|
|
DROP TABLE IF EXISTS objectEvent_v2;
|