From e795dfbd985d0239529edfa2de8f5aed0d59856a Mon Sep 17 00:00:00 2001 From: Oyvind Albrigtsen Date: Wed, 3 May 2023 12:32:14 +0200 Subject: [PATCH] - fence_aws: add --skip-race-check parameter to allow running outside of AWS network Resolves: rhbz#2183158 --- .gitignore | 5 +- ..._aws-1-add-skip-race-check-parameter.patch | 165 ++++++++++++++++++ ...fail-when-power-action-request-fails.patch | 21 +++ fence-agents.spec | 11 +- 4 files changed, 199 insertions(+), 3 deletions(-) create mode 100644 bz2183158-fence_aws-1-add-skip-race-check-parameter.patch create mode 100644 bz2183158-fence_aws-2-fail-when-power-action-request-fails.patch diff --git a/.gitignore b/.gitignore index 2d7a581..a25db22 100644 --- a/.gitignore +++ b/.gitignore @@ -1,9 +1,10 @@ +/*.tar.?z* /*.rpm -/*.tar.?z +/*.txt /*.whl +/*.zip /.*.swp /.build-*.log -/*.txt /*/ !/tests/ /tests/*.retry diff --git a/bz2183158-fence_aws-1-add-skip-race-check-parameter.patch b/bz2183158-fence_aws-1-add-skip-race-check-parameter.patch new file mode 100644 index 0000000..254e4ff --- /dev/null +++ b/bz2183158-fence_aws-1-add-skip-race-check-parameter.patch @@ -0,0 +1,165 @@ +From 73fdae1b9da5aa1ba1d371dcc47fe31a4d22bb31 Mon Sep 17 00:00:00 2001 +From: Oyvind Albrigtsen +Date: Thu, 30 Mar 2023 12:20:05 +0200 +Subject: [PATCH] fence_aws: fixes to allow running outside of AWS network + +- add --skip-race-check parameter to allow running outside of AWS + network e.g. for openshift +- fixed and improved logging logic +- use --debug-file parameter for file logging +--- + agents/aws/fence_aws.py | 50 ++++++++++++++++++++----------- + tests/data/metadata/fence_aws.xml | 5 ++++ + 2 files changed, 37 insertions(+), 18 deletions(-) + +diff --git a/agents/aws/fence_aws.py b/agents/aws/fence_aws.py +index c947bf29c..5d1677144 100644 +--- a/agents/aws/fence_aws.py ++++ b/agents/aws/fence_aws.py +@@ -16,13 +16,13 @@ + except ImportError: + pass + +-logger = logging.getLogger("fence_aws") ++logger = logging.getLogger() + logger.propagate = False + logger.setLevel(logging.INFO) + logger.addHandler(SyslogLibHandler()) + logging.getLogger('botocore.vendored').propagate = False + +-def get_instance_id(): ++def get_instance_id(options): + try: + token = requests.put('http://169.254.169.254/latest/api/token', headers={"X-aws-ec2-metadata-token-ttl-seconds" : "21600"}).content.decode("UTF-8") + r = requests.get('http://169.254.169.254/latest/meta-data/instance-id', headers={"X-aws-ec2-metadata-token" : token}).content.decode("UTF-8") +@@ -30,12 +30,15 @@ def get_instance_id(): + except HTTPError as http_err: + logger.error('HTTP error occurred while trying to access EC2 metadata server: %s', http_err) + except Exception as err: +- logger.error('A fatal error occurred while trying to access EC2 metadata server: %s', err) ++ if "--skip-race-check" not in options: ++ logger.error('A fatal error occurred while trying to access EC2 metadata server: %s', err) ++ else: ++ logger.debug('A fatal error occurred while trying to access EC2 metadata server: %s', err) + return None +- ++ + + def get_nodes_list(conn, options): +- logger.info("Starting monitor operation") ++ logger.debug("Starting monitor operation") + result = {} + try: + if "--filter" in options: +@@ -63,7 +66,7 @@ def get_power_status(conn, options): + try: + instance = conn.instances.filter(Filters=[{"Name": "instance-id", "Values": [options["--plug"]]}]) + state = list(instance)[0].state["Name"] +- logger.info("Status operation for EC2 instance %s returned state: %s",options["--plug"],state.upper()) ++ logger.debug("Status operation for EC2 instance %s returned state: %s",options["--plug"],state.upper()) + if state == "running": + return "on" + elif state == "stopped": +@@ -78,7 +81,7 @@ def get_power_status(conn, options): + except IndexError: + fail(EC_STATUS) + except Exception as e: +- logging.error("Failed to get power status: %s", e) ++ logger.error("Failed to get power status: %s", e) + fail(EC_STATUS) + + def get_self_power_status(conn, instance_id): +@@ -86,10 +89,10 @@ def get_self_power_status(conn, instance_id): + instance = conn.instances.filter(Filters=[{"Name": "instance-id", "Values": [instance_id]}]) + state = list(instance)[0].state["Name"] + if state == "running": +- logging.debug("Captured my (%s) state and it %s - returning OK - Proceeding with fencing",instance_id,state.upper()) ++ logger.debug("Captured my (%s) state and it %s - returning OK - Proceeding with fencing",instance_id,state.upper()) + return "ok" + else: +- logging.debug("Captured my (%s) state it is %s - returning Alert - Unable to fence other nodes",instance_id,state.upper()) ++ logger.debug("Captured my (%s) state it is %s - returning Alert - Unable to fence other nodes",instance_id,state.upper()) + return "alert" + + except ClientError: +@@ -100,18 +103,18 @@ def get_self_power_status(conn, instance_id): + return "fail" + + def set_power_status(conn, options): +- my_instance = get_instance_id() ++ my_instance = get_instance_id(options) + try: + if (options["--action"]=="off"): +- if (get_self_power_status(conn,my_instance) == "ok"): ++ if "--skip-race-check" in options or get_self_power_status(conn,my_instance) == "ok": + conn.instances.filter(InstanceIds=[options["--plug"]]).stop(Force=True) +- logger.info("Called StopInstance API call for %s", options["--plug"]) ++ logger.debug("Called StopInstance API call for %s", options["--plug"]) + else: +- logger.info("Skipping fencing as instance is not in running status") ++ logger.debug("Skipping fencing as instance is not in running status") + elif (options["--action"]=="on"): + conn.instances.filter(InstanceIds=[options["--plug"]]).start() + except Exception as e: +- logger.error("Failed to power %s %s: %s", \ ++ logger.debug("Failed to power %s %s: %s", \ + options["--action"], options["--plug"], e) + + def define_new_opts(): +@@ -156,12 +159,20 @@ def define_new_opts(): + "default": "False", + "order": 6 + } ++ all_opt["skip_race_check"] = { ++ "getopt" : "", ++ "longopt" : "skip-race-check", ++ "help" : "--skip-race-check Skip race condition check", ++ "shortdesc": "Skip race condition check", ++ "required": "0", ++ "order": 7 ++ } + + # Main agent method + def main(): + conn = None + +- device_opt = ["port", "no_password", "region", "access_key", "secret_key", "filter", "boto3_debug"] ++ device_opt = ["port", "no_password", "region", "access_key", "secret_key", "filter", "boto3_debug", "skip_race_check"] + + atexit.register(atexit_handler) + +@@ -183,12 +194,15 @@ def main(): + + run_delay(options) + +- if options.get("--verbose") is not None: +- lh = logging.FileHandler('/var/log/fence_aws_debug.log') ++ if "--debug-file" in options: ++ for handler in logger.handlers: ++ if isinstance(handler, logging.FileHandler): ++ logger.removeHandler(handler) ++ lh = logging.FileHandler(options["--debug-file"]) + logger.addHandler(lh) + lhf = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s') + lh.setFormatter(lhf) +- logger.setLevel(logging.DEBUG) ++ lh.setLevel(logging.DEBUG) + + if options["--boto3_debug"].lower() not in ["1", "yes", "on", "true"]: + boto3.set_stream_logger('boto3',logging.INFO) +diff --git a/tests/data/metadata/fence_aws.xml b/tests/data/metadata/fence_aws.xml +index 76995ecf2..32de4418a 100644 +--- a/tests/data/metadata/fence_aws.xml ++++ b/tests/data/metadata/fence_aws.xml +@@ -46,6 +46,11 @@ For instructions see: https://boto3.readthedocs.io/en/latest/guide/quickstart.ht + + Boto Lib debug + ++ ++ ++ ++ Skip race condition check ++ + + + diff --git a/bz2183158-fence_aws-2-fail-when-power-action-request-fails.patch b/bz2183158-fence_aws-2-fail-when-power-action-request-fails.patch new file mode 100644 index 0000000..63ebe1a --- /dev/null +++ b/bz2183158-fence_aws-2-fail-when-power-action-request-fails.patch @@ -0,0 +1,21 @@ +From a2e2432cfec0af9a8a90f9d7fed18759da6f9b0c Mon Sep 17 00:00:00 2001 +From: Oyvind Albrigtsen +Date: Thu, 13 Apr 2023 10:14:31 +0200 +Subject: [PATCH] fence_aws: fail when power action request fails + +--- + agents/aws/fence_aws.py | 1 + + 1 file changed, 1 insertion(+) + +diff --git a/agents/aws/fence_aws.py b/agents/aws/fence_aws.py +index 5d1677144..0a375bbec 100644 +--- a/agents/aws/fence_aws.py ++++ b/agents/aws/fence_aws.py +@@ -116,6 +116,7 @@ def set_power_status(conn, options): + except Exception as e: + logger.debug("Failed to power %s %s: %s", \ + options["--action"], options["--plug"], e) ++ fail(EC_STATUS) + + def define_new_opts(): + all_opt["region"] = { diff --git a/fence-agents.spec b/fence-agents.spec index 06ea06d..9960ea4 100644 --- a/fence-agents.spec +++ b/fence-agents.spec @@ -87,7 +87,7 @@ Name: fence-agents Summary: Set of unified programs capable of host isolation ("fencing") Version: 4.2.1 -Release: 113%{?alphatag:.%{alphatag}}%{?dist} +Release: 114%{?alphatag:.%{alphatag}}%{?dist} License: GPLv2+ and LGPLv2+ Group: System Environment/Base URL: https://github.com/ClusterLabs/fence-agents @@ -268,6 +268,8 @@ Patch125: bz2160478-fence_scsi-fix-validate-all.patch Patch126: bz2152105-fencing-1-add-plug_separator.patch Patch127: bz2152105-fencing-2-update-DEPENDENCY_OPT.patch Patch128: bz2187329-fence_scsi-detect-devices-in-shared-vgs.patch +Patch129: bz2183158-fence_aws-1-add-skip-race-check-parameter.patch +Patch130: bz2183158-fence_aws-2-fail-when-power-action-request-fails.patch %if 0%{?fedora} || 0%{?rhel} > 7 %global supportedagents amt_ws apc apc_snmp bladecenter brocade cisco_mds cisco_ucs compute drac5 eaton_snmp emerson eps evacuate hds_cb hpblade ibmblade ibm_powervs ibm_vpc ifmib ilo ilo_moonshot ilo_mp ilo_ssh intelmodular ipdu ipmilan kdump kubevirt lpar mpath redfish rhevm rsa rsb sbd scsi vmware_rest vmware_soap wti @@ -474,6 +476,8 @@ BuildRequires: python3-google-api-client python3-pip python3-wheel python3-jinja %patch126 -p1 %patch127 -p1 %patch128 -p1 +%patch129 -p1 -F2 +%patch130 -p1 # prevent compilation of something that won't get used anyway sed -i.orig 's|FENCE_ZVM=1|FENCE_ZVM=0|' configure.ac @@ -1471,6 +1475,11 @@ Fence agent for IBM z/VM over IP. %endif %changelog +* Wed May 3 2023 Oyvind Albrigtsen - 4.2.1-114 +- fence_aws: add --skip-race-check parameter to allow running outside + of AWS network + Resolves: rhbz#2183158 + * Thu Apr 20 2023 Oyvind Albrigtsen - 4.2.1-113 - fence_scsi: detect devices in shared VGs Resolves: rhbz#2187329