Merge remote-tracking branch 'origin/deprecate-databricks-job-name' i…

…nto test/gprofiler-no-missing-granulate-utils-dep

README.md

@@ -197,11 +197,6 @@ Additionally, 2 more flags need to be added to gProfiler's commandline: `--disab
 * `--disable-pidns-check` is required because gProfiler won't run in the init PID NS.
 * `--perf-mode=none` is required because gProfiler will not have permissions to run system-wide `perf`, so we will profile only runtime processes, such as Java. See [perf-less mode](#perf-less-mode) for more information.
 
-### Databricks unique service names for job clusters
-By using `--databricks-job-name-as-service-name`, gProfiler will use the Job Clusters' Job Name as service name.
-In case gProfiler successfully managed to extract the Job Name, the service name will be `databricks-job-<JOB_NAME>`.
-By default, this functionality relies on `spark.databricks.clusterUsageTags.clusterAllTags` property
-to extract the Job Name.
 
 In case gProfiler spots this property is redacted, gProfiler will use the
 `spark.databricks.clusterUsageTags.clusterName` property as service name.

executable.Dockerfile

@@ -74,8 +74,10 @@ RUN ./phpspy_build.sh
 # async-profiler glibc
 FROM centos${AP_BUILDER_CENTOS} AS async-profiler-builder-glibc
 WORKDIR /tmp
-
-COPY scripts/async_profiler_env_glibc.sh .
+COPY scripts/async_profiler_env_glibc.sh scripts/fix_centos7.sh ./
+RUN if grep -q "CentOS Linux 7" /etc/os-release ; then \
+      ./fix_centos7.sh; \
+    fi
 RUN ./async_profiler_env_glibc.sh
 
 COPY scripts/async_profiler_build_shared.sh .
@@ -139,10 +141,12 @@ RUN ./pyperf_build.sh --with-staticx
 FROM centos${GPROFILER_BUILDER} AS build-prepare
 
 WORKDIR /tmp
-COPY scripts/fix_centos8.sh .
+COPY scripts/fix_centos7.sh scripts/fix_centos8.sh ./
 # fix repo links for CentOS 8, and enable powertools (required to download glibc-static)
 RUN if grep -q "CentOS Linux 8" /etc/os-release ; then \
         ./fix_centos8.sh; \
+    elif grep -q "CentOS Linux 7" /etc/os-release ; then \
+        ./fix_centos7.sh; \
     fi
 
 # update libmodulemd to fix https://bugzilla.redhat.com/show_bug.cgi?id=2004853
@@ -200,7 +204,10 @@ RUN python3 -m pip install --no-cache-dir --upgrade pip
 FROM ${NODE_PACKAGE_BUILDER_GLIBC} as node-package-builder-glibc
 USER 0
 WORKDIR /tmp
-COPY scripts/node_builder_glibc_env.sh .
+COPY scripts/node_builder_glibc_env.sh scripts/fix_centos7.sh ./
+RUN if grep -q "CentOS Linux 7" /etc/os-release ; then \
+      ./fix_centos7.sh; \
+    fi
 RUN ./node_builder_glibc_env.sh
 COPY scripts/build_node_package.sh .
 RUN ./build_node_package.sh

gprofiler/main.py

@@ -34,7 +34,6 @@
 from granulate_utils.linux.ns import is_running_in_init_pid
 from granulate_utils.linux.process import is_process_running
 from granulate_utils.metadata.cloud import get_aws_execution_env
-from granulate_utils.metadata.databricks_client import DBXWebUIEnvWrapper, get_name_from_metadata
 from psutil import NoSuchProcess, Process
 from requests import RequestException, Timeout
 
@@ -798,11 +797,7 @@ def parse_cmd_args() -> configargparse.Namespace:
         action="store_true",
         dest="databricks_job_name_as_service_name",
         default=False,
-        help="gProfiler will set service name to Databricks' job name on ephemeral clusters. It'll delay the beginning"
-        " of the profiling due to repeated waiting for Spark's metrics server."
-        ' service name format is: "databricks-job-<JOB-NAME>".'
-        " Note that in any case that the job name is not available due to redaction,"
-        " gProfiler will fallback to use the clusterName property.",
+        help="Deprecated! Removed in version 1.49.0",
     )
 
     parser.add_argument(
@@ -1019,6 +1014,9 @@ def warn_about_deprecated_args(args: configargparse.Namespace) -> None:
     if args.collect_spark_metrics:
         logger.warning("--collect-spark-metrics is deprecated and removed in version 1.42.0")
 
+    if args.databricks_job_name_as_service_name:
+        logger.warning("--databricks-job-name-as-service-name is deprecated and removed in version 1.49.0")
+
 
 def main() -> None:
     args = parse_cmd_args()
@@ -1059,19 +1057,6 @@ def main() -> None:
     # assume we run in the root cgroup (when containerized, that's our view)
     usage_logger = CgroupsUsageLogger(logger, "/") if args.log_usage else NoopUsageLogger()
 
-    if args.databricks_job_name_as_service_name:
-        # "databricks" will be the default name in case of failure with --databricks-job-name-as-service-name flag
-        args.service_name = "databricks"
-        dbx_web_ui_wrapper = DBXWebUIEnvWrapper(logger)
-        dbx_metadata = dbx_web_ui_wrapper.all_props_dict
-        if dbx_metadata is not None:
-            service_suffix = get_name_from_metadata(dbx_metadata)
-            if service_suffix is not None:
-                args.service_name = f"databricks-{service_suffix}"
-
-        if remote_logs_handler is not None:
-            remote_logs_handler.update_service_name(args.service_name)
-
     try:
         logger.info(
             "Running gProfiler", version=__version__, commandline=" ".join(sys.argv[1:]), arguments=args.__dict__

scripts/fix_centos7.sh

@@ -0,0 +1,8 @@
+#!/usr/bin/env bash
+
+set -eu
+
+sed -i 's/mirror.centos.org/vault.centos.org/g' /etc/yum.repos.d/*.repo
+sed -i 's/^#.*baseurl=http/baseurl=http/g' /etc/yum.repos.d/*.repo
+sed -i 's/^mirrorlist=http/#mirrorlist=http/g' /etc/yum.repos.d/*.repo
+yum clean all