Fixed execute_init_commands bug on L446 (#368)

Author: filipecosta90

poetry.lock

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "redisbench-admin"
-version = "0.9.9"
+version = "0.9.12"
 description = "Redis benchmark run helper. A wrapper around Redis and Redis Modules benchmark tools ( ftsb_redisearch, memtier_benchmark, redis-benchmark, aibench, etc... )."
 authors = ["filipecosta90 <[email protected]>","Redis Performance Group <[email protected]>"]
 readme = "README.md"
@@ -10,7 +10,7 @@ redisbench-admin = "redisbench_admin.cli:main"
 perf-daemon = "redisbench_admin.profilers.daemon:main"
 
 [tool.poetry.dependencies]
-python = "^3.6.1"
+python = "^3.6.9"
 humanize = "^2.4.0"
 requests = "^2.23.0"
 py_cpuinfo = "^5.0.0"
@@ -22,7 +22,7 @@ jsonpath_ng = "^1.5.2"
 pysftp = "^0.2.9"
 python_terraform = "^0.10.1"
 GitPython = "^3.1.12"
-PyYAML = "^5.4"
+PyYAML = "^6.0"
 wget = "^3.2"
 pytablewriter = {extras = ["html"], version = "^0.64.1"}
 sshtunnel = "^0.4.0"

pyproject.toml

@@ -64,7 +64,7 @@ def create_export_arguments(parser):
         type=str,
         default="json",
         help="results format of the the benchmark results files to read "
-        "results from ( either pyperf-json, csv, json, redis-benchmark-txt )",
+        "results from ( either google.benchmark, pyperf-json, csv, json, redis-benchmark-txt )",
     )
     parser.add_argument(
         "--use-result",

redisbench_admin/export/args.py

@@ -12,6 +12,9 @@
 
 
 from redisbench_admin.export.common.common import split_tags_string
+from redisbench_admin.export.google_benchmark.google_benchmark_json_format import (
+    generate_summary_json_google_benchmark,
+)
 from redisbench_admin.export.pyperf.pyperf_json_format import (
     generate_summary_json_pyperf,
 )
@@ -50,10 +53,13 @@ def export_command_logic(args, project_name, project_version):
             "You need to specify at least one (or more) of --deployment-version --github_branch arguments"
         )
         exit(1)
-    if results_format != "csv" and results_format != "pyperf-json":
+    non_required_spec = ["csv", "pyperf-json", "google.benchmark"]
+    if results_format not in non_required_spec:
         if exporter_spec_file is None:
             logging.error(
-                "--exporter-spec-file is required for all formats with exception of csv and pyperf-json"
+                "--exporter-spec-file is required for all formats with exception of {}".format(
+                    ",".join(non_required_spec)
+                )
             )
             exit(1)
         else:
@@ -76,6 +82,22 @@ def export_command_logic(args, project_name, project_version):
         with open(benchmark_file, "r") as json_file:
             start_dict = json.load(json_file)
             results_dict = generate_summary_json_pyperf(start_dict)
+    if results_format == "google.benchmark":
+        with open(benchmark_file, "r") as json_file:
+            # override test names
+            print_warning = False
+            old_test_name = test_name
+            if test_name is None:
+                print_warning = True
+            start_dict = json.load(json_file)
+            results_dict, test_name = generate_summary_json_google_benchmark(start_dict)
+            if print_warning is True:
+                logging.warning(
+                    "You've specificied a test name {} but on google benchmark we override it based on the test names retrieved from out file {}".format(
+                        old_test_name, test_name
+                    )
+                )
+
     if args.override_test_time:
         datapoints_timestamp = int(args.override_test_time.timestamp() * 1000.0)
         logging.info(
@@ -120,9 +142,9 @@ def export_command_logic(args, project_name, project_version):
             triggering_env,
         )
         logging.info("Parsed a total of {} metrics".format(len(timeseries_dict.keys())))
-    if results_format == "pyperf-json":
-        logging.info("Parsing pyperf format into timeseries format")
-        timeseries_dict = export_pyperf_json_to_timeseries_dict(
+    if results_format == "pyperf-json" or results_format == "google.benchmark":
+        logging.info("Parsing {} format into timeseries format".format(results_format))
+        timeseries_dict = export_json_to_timeseries_dict(
             results_dict,
             break_by_dict,
             datapoints_timestamp,
@@ -181,7 +203,7 @@ def export_command_logic(args, project_name, project_version):
     )
 
 
-def export_pyperf_json_to_timeseries_dict(
+def export_json_to_timeseries_dict(
     benchmark_file,
     break_by_dict,
     datapoints_timestamp,

redisbench_admin/export/export.py

@@ -0,0 +1,5 @@
+#  Apache License Version 2.0
+#
+#  Copyright (c) 2021., Redis Labs Modules
+#  All rights reserved.
+#

redisbench_admin/export/google_benchmark/__init__.py

@@ -0,0 +1,53 @@
+#  Apache License Version 2.0
+#
+#  Copyright (c) 2021., Redis Labs Modules
+#  All rights reserved.
+#
+import logging
+
+import numpy as np
+
+
+def metric_safe_name(row, replace_by="_"):
+    import re
+
+    metric_name = row.strip()
+    metric_name = re.sub(r"\W+", replace_by, metric_name)
+    return metric_name
+
+
+def generate_summary_json_google_benchmark(input_json):
+    result_json = {}
+    test_names = []
+    for benchmark in input_json["benchmarks"]:
+        original_name = benchmark["name"]
+        benchmark_name = original_name
+        non_safe_count = len(original_name) - len(metric_safe_name(original_name, ""))
+        if non_safe_count > 0:
+            benchmark_name = metric_safe_name(original_name)
+            while "_" == benchmark_name[len(benchmark_name) - 1]:
+                benchmark_name = benchmark_name[: len(benchmark_name) - 1]
+            logging.warning(
+                "Given the benchmark name {} contains {} non alphanumeric characters, we're replacing it by the safe version {}".format(
+                    original_name, "-", benchmark_name
+                )
+            )
+        metrics = {}
+        test_names.append(benchmark_name)
+        for metric_name, metric_value_str in benchmark.items():
+            metric_value = None
+            try:
+                metric_value = float(metric_value_str)
+            except ValueError:
+                pass
+            if metric_value is not None:
+                logging.info(
+                    "Adding google.benchmark to benchmark {} metric named {}={}".format(
+                        benchmark_name, metric_name, metric_value
+                    )
+                )
+                metrics[metric_name] = metric_value
+
+        result_json[benchmark_name] = metrics
+
+    return result_json, test_names

redisbench_admin/export/google_benchmark/google_benchmark_json_format.py

@@ -63,7 +63,7 @@ def common_run_args(parser):
     parser.add_argument(
         "--allowed-tools",
         type=str,
-        default="memtier_benchmark,redis-benchmark,redisgraph-benchmark-go,ycsb,"
+        default="memtier_benchmark,redis-benchmark,redisgraph-benchmark-go,ycsb,go-ycsb,"
         + "tsbs_run_queries_redistimeseries,tsbs_load_redistimeseries,"
         + "ftsb_redisearch,"
         + "aibench_run_inference_redisai_vision,ann-benchmarks",

redisbench_admin/run/args.py

@@ -33,7 +33,10 @@
 from redisbench_admin.run.tsbs_run_queries_redistimeseries.tsbs_run_queries_redistimeseries import (
     prepare_tsbs_benchmark_command,
 )
-from redisbench_admin.run.ycsb.ycsb import prepare_ycsb_benchmark_command
+from redisbench_admin.run.ycsb.ycsb import (
+    prepare_ycsb_benchmark_command,
+    prepare_go_ycsb_benchmark_command,
+)
 from redisbench_admin.run_remote.remote_helpers import (
     extract_module_semver_from_info_modules_cmd,
 )
@@ -168,7 +171,7 @@ def prepare_benchmark_parameters_specif_tooling(
             remote_results_file,
             isremote,
         )
-    if "ycsb" in benchmark_tool:
+    if "ycsb" in benchmark_tool and "go-ycsb" not in benchmark_tool:
         if isremote is True:
             benchmark_tool = "/tmp/ycsb/bin/ycsb"
             current_workdir = "/tmp/ycsb"
@@ -179,6 +182,18 @@ def prepare_benchmark_parameters_specif_tooling(
             entry,
             current_workdir,
         )
+    if "go-ycsb" in benchmark_tool:
+        if isremote is True:
+            benchmark_tool = "/tmp/{}".format(benchmark_tool)
+        command_arr, command_str = prepare_go_ycsb_benchmark_command(
+            benchmark_tool,
+            server_private_ip,
+            server_plaintext_port,
+            entry,
+            current_workdir,
+            cluster_api_enabled,
+        )
+
     if "tsbs_" in benchmark_tool:
         input_data_file = None
         if isremote is True:
@@ -417,10 +432,25 @@ def execute_init_commands(benchmark_config, r, dbconfig_keyname="dbconfig"):
                 is_array = True
             try:
                 logging.info("Sending init command: {}".format(cmd))
+                stdout = ""
                 if is_array:
-                    stdout = r.execute_command(*cmd)
+                    if "FT.CREATE" in cmd[0]:
+                        logging.info("Detected FT.CREATE to all nodes on OSS Cluster")
+                        try:
+                            stdout = r.execute_command(*cmd, target_nodes="all")
+                        except redis.exceptions.ResponseError:
+                            pass
+                    else:
+                        stdout = r.execute_command(*cmd)
                 else:
-                    stdout = r.execute_command(cmd)
+                    if "FT.CREATE" in cmd:
+                        logging.info("Detected FT.CREATE to all nodes on OSS Cluster")
+                        try:
+                            stdout = r.execute_command(cmd, target_nodes="all")
+                        except redis.exceptions.ResponseError:
+                            pass
+                    else:
+                        stdout = r.execute_command(cmd)
                 logging.info("Command reply: {}".format(stdout))
             except redis.connection.ConnectionError as e:
                 logging.error(

redisbench_admin/run/common.py

@@ -315,25 +315,65 @@ def timeseries_test_sucess_flow(
                 rts.hset(
                     branch_target_table_keyname, None, None, branch_target_table_dict
                 )
-
-        update_secondary_result_keys(
-            artifact_version,
-            benchmark_duration_seconds,
-            build_variant_name,
-            dataset_load_duration_seconds,
-            deployment_name,
-            deployment_type,
-            metadata_tags,
-            rts,
-            running_platform,
-            start_time_ms,
-            test_name,
-            testcase_metric_context_paths,
-            tf_github_branch,
-            tf_github_org,
-            tf_github_repo,
-            tf_triggering_env,
-        )
+        if test_name is not None:
+            if type(test_name) is str:
+                update_secondary_result_keys(
+                    artifact_version,
+                    benchmark_duration_seconds,
+                    build_variant_name,
+                    dataset_load_duration_seconds,
+                    deployment_name,
+                    deployment_type,
+                    metadata_tags,
+                    rts,
+                    running_platform,
+                    start_time_ms,
+                    test_name,
+                    testcase_metric_context_paths,
+                    tf_github_branch,
+                    tf_github_org,
+                    tf_github_repo,
+                    tf_triggering_env,
+                )
+            if type(test_name) is list:
+                for inner_test_name in test_name:
+                    update_secondary_result_keys(
+                        artifact_version,
+                        benchmark_duration_seconds,
+                        build_variant_name,
+                        dataset_load_duration_seconds,
+                        deployment_name,
+                        deployment_type,
+                        metadata_tags,
+                        rts,
+                        running_platform,
+                        start_time_ms,
+                        inner_test_name,
+                        testcase_metric_context_paths,
+                        tf_github_branch,
+                        tf_github_org,
+                        tf_github_repo,
+                        tf_triggering_env,
+                    )
+        else:
+            update_secondary_result_keys(
+                artifact_version,
+                benchmark_duration_seconds,
+                build_variant_name,
+                dataset_load_duration_seconds,
+                deployment_name,
+                deployment_type,
+                metadata_tags,
+                rts,
+                running_platform,
+                start_time_ms,
+                test_name,
+                testcase_metric_context_paths,
+                tf_github_branch,
+                tf_github_org,
+                tf_github_repo,
+                tf_triggering_env,
+            )
     return version_target_tables, branch_target_tables
 
 

redisbench_admin/run/redistimeseries.py

@@ -52,17 +52,21 @@ def prepare_tsbs_benchmark_command(
     return command_arr, command_str
 
 
-def extract_tsbs_extra_links(
+def extract_remote_tool_extra_links(
     benchmark_config,
     benchmark_tool,
     config_key="clientconfig",
     os_str="linux",
     arch_str="amd64",
+    project="redistimeseries",
+    tools_group="tsbs",
 ):
     remote_tool_link = "/tmp/{}".format(benchmark_tool)
     tool_link = (
         "https://s3.amazonaws.com/benchmarks.redislabs/"
-        + "redistimeseries/tools/tsbs/{}_{}_{}".format(benchmark_tool, os_str, arch_str)
+        + "{}/tools/{}/{}_{}_{}".format(
+            project, tools_group, benchmark_tool, os_str, arch_str
+        )
     )
     queries_file_link = None
     for entry in benchmark_config[config_key]: