Skip to main content

Apache Hive

Connect Apache Hive so NudgeBee can query logs stored in a Hive table. Hive is a log-only source.

caution

For this integration to work correctly, the Hive table must be partitioned by time (year / month / day / hour), and your filter must include at least one of those partition columns. Without partition pruning, Hive scans the entire table — queries are slow and may fail outright on malformed rows.

Prerequisites​

  • A HiveServer2 endpoint reachable from NudgeBee, in host:port form, e.g. hiveserver2.hive.svc.cluster.local:10000.
  • A Hive table holding log data, partitioned by time.
  • The column names for the log timestamp and the log message body.

Step 1: Configure the Integration in NudgeBee​

Navigate to Admin > Integrations > Observability, select Apache Hive, then click Add Hive Account.

The Apache Hive tile in the Observability tab of the integrations catalog

  • Integration Config Name * (Required) — a name for this Hive integration, e.g. app-logs-hive.
  • Account ID * (Required) — the NudgeBee account to link.
  • Hive URL * (Required) — the HiveServer2 endpoint as host and port, e.g. hiveserver2.hive.svc:10000.
  • Hive Database — the database to connect to. Defaults to default.
  • Auth Type — the authentication method. Defaults to None.
  • Hive Table * (Required) — the table containing log data, e.g. logs.app_logs.
  • Hive Timestamp Col * (Required) — the column storing the log timestamp, as a numeric epoch or a formatted string.
  • The column storing the log message body is also required.

The Apache Hive configuration form, showing the HiveServer2 endpoint, table, and timestamp and message column settings

Step 2: Test and Save​

Click Test Connection, then Save.

Verify the Integration​

  1. Click Test Connection on the form — it should succeed before you save.
  2. Open Clusters, select a cluster, and go to Monitoring > Logs > Query Log.
  3. Confirm log rows from Hive appear. Include a time range so partition pruning applies.

Troubleshooting​

SymptomLikely CauseFix
Queries are very slow or time outThe filter does not include a partition columnEnsure the table is partitioned by time and that queries include the partition column.
Query fails on some rowsMalformed rows in the tableHive scans without pruning will surface bad rows; fix the source data or narrow the partition range.
Connection refusedWrong HiveServer2 portHiveServer2 usually listens on 10000; confirm the port your deployment uses.
No results despite data existingWrong timestamp column, or a format mismatchConfirm Hive Timestamp Col and whether it stores epoch values or formatted strings.
Authentication errorsAuth Type does not match the serverSet the authentication method your HiveServer2 requires.