Event storage structure¶

Overview¶

You can see how transmitted event logs are stored in BigQuery, the detailed structure of the attributes and JSON fields, and how logs that violate transmission rules are quarantined.

table structure¶

Sent event logs are stored in the t_raw_event_log table in the company-specific BigQuery dataset.

Table column structure¶

Storage location¶

Column definition¶

attributes example

{
  "hiveAttributes": {
    "dataSource": "custom_server",
    "geoIpCountry": "KR"
  },
  "eventAttributes": {
    "eventTime": "2026-07-20T14:01:01+09:00",
    "level": 10,
    "character_name": "AA",
    "stage_id": "stage_101"
  }
}

Table structure precautions & Tips¶

• The same event may be stored redundantly due to retransmission, etc., so it is recommended to use a duplicate removal query based on checksum when searching.
• Event logs that violate transport rules are stored in the rescue table, not this table. For more information, see Data Isolation.

attributes Detailed attributes¶

attributes example¶

{
  "hiveAttributes": {
    "dataSource": "custom_server",
    "geoIpCountry": "KR"
  },
  "eventAttributes": {
    "eventTime": "2026-07-20T14:01:01+09:00",
    "level": 10,
    "character_name": "AA"
  }
}

hiveAttributes¶

This is the area that the pipeline automatically processes and stores.

dataSource Details¶

dataSource indicates the entity sending the event log, and is automatically set to one of the following values:

eventAttributes¶

This is the area where the attributes sent by the user and the attributes automatically collected by SDK are stored.

attributes Precautions & Tips¶

• hiveAttributes is not an area where the developer sets the value directly, but an area that the pipeline automatically fills. Values ​​other than dataSource and geoIpCountry are not saved.
• Both custom attributes and SDK auto-collected attributes are stored together in eventAttributes. When searching, they must be distinguished by attribute name.
• You can search by distinguishing the transmission path of the event log (SDK/ server/MMP, etc.) using the dataSource value.

data-isolation¶

Reason for quarantine¶

Isolated table structure¶

Isolated event logs are stored in the t_raw_event_log_rescue table with the following structure:

Related menu¶

• Event attribute transmission — Detailed rules for required/automatically collected/custom-defined attributes
• Event log transmission — Transmission method via Hive SDK v4, Hive Axyl, Fluentd, HTTP