Add vllm_buildkite_agents table (#7365)

The PR to create the new table on test-infra side. I have test the query
on `fortesting` table to start with.

cc @hl475

---------

Signed-off-by: Huy Do <[email protected]>

Author: huydhn

aws/lambda/clickhouse-replicator-dynamo/lambda_function.py

@@ -30,6 +30,7 @@
     "torchci-pull-request-review-comment": "default.pull_request_review_comment",
     "torchci-metrics-ci-wait-time": "misc.metrics_ci_wait_time",
     "torchci-dynamo-perf-stats": "benchmark.inductor_torch_dynamo_perf_stats",
+    "vllm-buildkite-agent-events": "vllm.vllm_buildkite_agents",
     "vllm-buildkite-build-events": "vllm.vllm_buildkite_builds",
     "vllm-buildkite-job-events": "vllm.vllm_buildkite_jobs",
 }

clickhouse_db_schema/vllm/vllm_buildkite_agents/schema.sql

@@ -0,0 +1,70 @@
+CREATE TABLE vllm.vllm_buildkite_agents (
+    dynamoKey String,
+    agent Tuple(
+        last_job_finished_at Nullable(DateTime64(3)),
+        creator Nullable(String),
+        paused Bool,
+        created_at DateTime64(3),
+        ip_address String,
+        connection_state String,
+        priority UInt32,
+        version String,
+        url String,
+        cluster_url String,
+        hostname String,
+        paused_at Nullable(DateTime64(3)),
+        web_url String,
+        cluster_queue_url String,
+        name String,
+        meta_data Array(String),
+        paused_timeout_in_minutes UInt32,
+        id String,
+        paused_note Nullable(String),
+        job Tuple(
+            retried Bool,
+            finished_at Nullable(DateTime64(3)),
+            parallel_group_index Nullable(UInt32),
+            graphql_id String,
+            artifacts_url String,
+            step_key Nullable(String),
+            created_at DateTime64(3),
+            scheduled_at DateTime64(3),
+            expired_at Nullable(DateTime64(3)),
+            retries_count Nullable(UInt32),
+            type String,
+            exit_status Nullable(Int32),
+            matrix Nullable(String),
+            cluster_url String,
+            cluster_id String,
+            soft_failed Bool,
+            retried_in_job_id Nullable(String),
+            log_url String,
+            retry_source Nullable(String),
+            id String,
+            state String,
+            build_url String,
+            retry_type Nullable(String),
+            priority Tuple(number UInt32),
+            agent_query_rules Array(String),
+            command String,
+            raw_log_url String,
+            cluster_queue_id String,
+            web_url String,
+            cluster_queue_url String,
+            parallel_group_total Nullable(UInt32),
+            name String,
+            started_at Nullable(DateTime64(3)),
+            step Tuple(signature Nullable(String), id String),
+            artifact_paths Nullable(String),
+            runnable_at DateTime64(3)
+        ),
+        user_agent String,
+        paused_by Nullable(String)
+    ),
+    event String,
+    sender Tuple(name Nullable(String), id Nullable(String))
+) ENGINE = SharedReplacingMergeTree('/clickhouse/tables/{uuid}/{shard}', '{replica}')
+ORDER BY (
+    agent.created_at,
+    agent.name
+) SETTINGS index_granularity = 8192