TRIVIAL: gooddata-dbt - better report stats from db cloud

pretty tables
report always, just change the prefix based on if there are any degradations

gooddata-dbt/gooddata_dbt/dbt_plugin.py

@@ -7,9 +7,10 @@
 from time import time
 from typing import List, Optional
 
+import tabulate
 import yaml
 from gooddata_dbt.args import parse_arguments
-from gooddata_dbt.dbt.cloud import DbtConnection, DbtCredentials
+from gooddata_dbt.dbt.cloud import DbtConnection, DbtCredentials, DbtExecution
 from gooddata_dbt.dbt.profiles import DbtOutput, DbtProfiles
 from gooddata_dbt.dbt.tables import DbtModelTables
 from gooddata_dbt.gooddata.config import GoodDataConfig, GoodDataConfigProduct
@@ -172,6 +173,7 @@ def deploy_models(
     dbt_target: DbtOutput,
     gd_config: GoodDataConfig,
     sdk: GoodDataSdk,
+    args: Namespace,
     data_source_id: str,
 ) -> None:
     dbt_tables = DbtModelTables.from_local(gooddata_upper_case, gd_config.all_model_ids)
@@ -187,11 +189,69 @@ def deploy_models(
                 workspace_id = f"{data_product.id}_{environment.id}"
                 workspace_title = f"{data_product.name} ({environment.name})"
                 create_workspace(logger, sdk, workspace_id, workspace_title)
-                deploy_ldm(logger, sdk, data_source_id, dbt_tables, data_product.model_ids, workspace_id)
+                deploy_ldm(logger, sdk, args, data_source_id, dbt_tables, data_product.model_ids, workspace_id)
                 if data_product.localization:
                     create_localized_workspaces(data_product, sdk, workspace_id)
 
 
+def get_table(data: list[list], headers: list[str], fmt: str) -> str:
+    return tabulate.tabulate(data, headers=headers, tablefmt=fmt)
+
+
+def dbt_cloud_stats_degradations(
+    args: Namespace,
+    logger: logging.Logger,
+    environment_id: str,
+    model_executions: List[DbtExecution],
+    dbt_conn: DbtConnection,
+):
+    logger.info("Get stats for historical executions...")
+    models_history_avg_execution_times = dbt_conn.get_average_times(logger, model_executions, environment_id, 5)
+    differences = []
+    degradations = 0
+    for execution in model_executions:
+        for model_id, avg_time in models_history_avg_execution_times.items():
+            last_execution = execution.execution_info.execution_time
+            difference = float(last_execution - avg_time) / avg_time * 100
+            degradation = difference > args.allowed_degradation
+            if model_id == execution.unique_id:
+                differences.append(
+                    [
+                        model_id,
+                        round(last_execution, 2),
+                        round(avg_time, 2),
+                        round(difference, 2),
+                        str(degradation),
+                    ]
+                )
+                if degradation:
+                    degradations += 1
+    headers = ["Model ID", "Last duration(s)", "Average duration(s)", "Difference(%)", "Degradation"]
+    pretty_table = get_table(differences, headers, "outline")
+    if degradations > 0:
+        logger.warning(
+            "Stats for historical executions contain degradations! "
+            + f"Threshold={args.allowed_degradation}%\n{pretty_table}"
+        )
+    else:
+        logger.info(f"Stats for historical executions:\n{pretty_table}")
+
+    gitlab_token = os.getenv("GITLAB_TOKEN")
+    if os.getenv("CI_MERGE_REQUEST_IID") and gitlab_token:
+        # Running in Gitlab CI pipeline, report performance of executions to the merge request to notify code reviewer
+        git_table = get_table(differences, headers, "github")
+        logger.info("Sending report to related Gitlab merge request as comment")
+        if degradations > 0:
+            message = (
+                "WARNING: some executions of dbt models in this merge request are slower than average!"
+                + f"Threshold={args.allowed_degradation}%\n\n{git_table}"
+            )
+            report_message_to_merge_request(gitlab_token, message)
+        else:
+            message = f"INFO: performance of all executions of dbt models is OK!\n\n{git_table}"
+            report_message_to_merge_request(gitlab_token, message)
+
+
 def dbt_cloud_stats(
     args: Namespace,
     logger: logging.Logger,
@@ -202,47 +262,15 @@ def dbt_cloud_stats(
     dbt_conn = DbtConnection(credentials=DbtCredentials(account_id=args.account_id, token=args.token))
     dbt_tables = DbtModelTables.from_local(args.gooddata_upper_case, all_model_ids)
     model_executions = dbt_conn.get_last_execution(environment_id, len(dbt_tables.tables))
+    exec_list = []
     for execution in model_executions:
-        logger.info(f"Model {execution.unique_id} finished in {execution.execution_info.execution_time:.2f}s")
+        exec_list.append([execution.unique_id, round(execution.execution_info.execution_time, 2)])
 
-    logger.info("Get stats for historical executions...")
-    models_history_avg_execution_times = dbt_conn.get_average_times(logger, model_executions, environment_id, 5)
-    degradations = {}
-    for execution in model_executions:
-        for model_id, avg_time in models_history_avg_execution_times.items():
-            degradation = float(execution.execution_info.execution_time - avg_time) / avg_time * 100
-            if model_id == execution.unique_id and degradation > args.allowed_degradation:
-                degradations[model_id] = {
-                    "last_time": execution.execution_info.execution_time,
-                    "avg_time": avg_time,
-                    "degradation": degradation,
-                }
-    degradation_md = (
-        "**WARNING:** The performance of some dbt cloud jobs degraded "
-        + f"by more than {args.allowed_degradation}%\n\n"
-        + "| Model | Last duration(s) | Average duration(s) | Degradation |\n"
-        + "|-------|------------------|---------------------|-------------|"
-    )
-    if len(degradations) > 0:
-        for model_id, data in degradations.items():
-            model_degradation_text = (
-                f"Model {model_id} - performance degraded: "
-                + f"last_time={data['last_time']:.2f} avg_time={data['avg_time']:.2f} "
-                + f"degradation={data['degradation']:.2f}%"
-            )
-            logger.warning(model_degradation_text)
-            model_degradation_md = "| {model_id} | {last_time} | {avg_time} | {degradation}% |".format(
-                model_id=model_id,
-                last_time=f"{data['last_time']:.2f}",
-                avg_time=f"{data['avg_time']:.2f}",
-                degradation=f"{data['degradation']:.2f}",
-            )
-            degradation_md += f"\n{model_degradation_md}"
-        gitlab_token = os.getenv("GITLAB_TOKEN")
-        if os.getenv("CI_MERGE_REQUEST_IID") and gitlab_token:
-            logger.info("Sending report of degradations to related Gitlab merge request as comment")
-            # Running in Gitlab CI pipeline, report degradations to the merge request to notify code reviewer
-            report_message_to_merge_request(gitlab_token, degradation_md)
+    headers = ["Model ID", "Duration(s)"]
+    pretty_table = get_table(exec_list, headers, "outline")
+    logger.info(f"Stats for last execution:\n{pretty_table}")
+
+    dbt_cloud_stats_degradations(args, logger, environment_id, model_executions, dbt_conn)
 
 
 def dbt_cloud_run(args: Namespace, logger: logging.Logger, all_model_ids: List[str]) -> None:
@@ -294,6 +322,7 @@ def main() -> None:
                 dbt_target,
                 gd_config,
                 sdk,
+                args,
                 data_source_id,
             )
     else:
@@ -306,7 +335,7 @@ def main() -> None:
                     if args.method == "store_analytics":
                         store_analytics(logger, sdk, workspace_id, data_product)
                     elif args.method == "deploy_analytics":
-                        deploy_analytics(logger, sdk, workspace_id, data_product)
+                        deploy_analytics(logger, sdk, args, workspace_id, data_product)
                     elif args.method == "test_insights":
                         test_insights(logger, sdk, workspace_id)
                     else:

gooddata-dbt/requirements.txt

@@ -2,3 +2,4 @@ pyyaml>=5.1
 attrs==21.4.0
 cattrs==22.1.0
 requests~=2.31.0
+tabulate~=0.8.10