def test_should_generate_res_count_graph_with_alert(self):
name = "service-1"
cloudwatch_data_source = "prod"
loadbalancer = "loadbalancer-1"
target_group = "target-group-1"
grid_pos = GridPos(1, 2, 3, 4)
notifications = ["foo", "bar", "baz"]
panel = generate_res_count_graph(
name=name,
cloudwatch_data_source=cloudwatch_data_source,
grid_pos=grid_pos,
loadbalancer=loadbalancer,
target_group=target_group,
notifications=notifications,
)
panel.alert.should.be.a(Alert)
panel.alert.message.should.eql("{} has 5XX errors".format(name))
panel.alert.alertConditions.should.have.length_of(1)
panel.alert.alertConditions.should.eql(
[
AlertCondition(
Target(refId="A"),
timeRange=TimeRange("15m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_AND,
),
]
)
def test_should_generate_pending_count_with_alerts_graph(self):
name = "service-1"
cloudwatch_data_source = "prod"
cluster_name = "cluster-1"
grid_pos = GridPos(1, 2, 3, 4)
notifications = ["foo", "bar"]
panel = generate_pending_count_graph(
name=name,
cloudwatch_data_source=cloudwatch_data_source,
cluster_name=cluster_name,
grid_pos=grid_pos,
notifications=notifications,
)
panel.alert.should.be.a(Alert)
panel.alert.gracePeriod.should.eql("15m")
panel.alert.alertConditions.should.have.length_of(1)
panel.alert.alertConditions[0].should.eql(
AlertCondition(
Target(refId="A"),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
)
def test_should_generate_mem_utilization_percentage_with_alerts_graph(self):
name = "service-1"
cloudwatch_data_source = "prod"
cluster_name = "cluster-1"
grid_pos = GridPos(1, 2, 3, 4)
notifications = ["foo", "bar", "baz"]
expected_alert_condition = AlertCondition(
Target(refId="A"),
timeRange=TimeRange("15m", "now"),
evaluator=GreaterThan(85),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
panel = generate_mem_utilization_percentage_graph(
name=name,
cloudwatch_data_source=cloudwatch_data_source,
cluster_name=cluster_name,
grid_pos=grid_pos,
notifications=notifications,
)
panel.alert.should.be.a(Alert)
panel.alert.alertConditions.should.have.length_of(1)
panel.alert.alertConditions[0].should.eql(expected_alert_condition)
panel.alert.notifications.should.eql(notifications)
def generate_elasticsearch_status_red_alert_graph(
name: str, client_id: str, cloudwatch_data_source: str,
notifications: List[str]) -> Graph:
"""
Generate Elasticsearch graph
"""
y_axes = YAxes(
YAxis(format=SHORT_FORMAT),
YAxis(format=SHORT_FORMAT),
)
targets = [
CloudwatchMetricsTarget(
alias="Red status",
namespace=NAMESPACE,
period="1m",
statistics=["Maximum"],
dimensions={
"DomainName": name,
"ClientId": client_id
},
metricName="ClusterStatus.red",
),
]
alert = None
if notifications:
alert = Alert(
name="Elasticsearch is in status red",
message="Elasticsearch is in status red",
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_OR,
),
],
frequency="2m",
gracePeriod="2m",
notifications=notifications,
)
return Graph(
title="Status RED alerts",
dataSource=cloudwatch_data_source,
targets=targets,
yAxes=y_axes,
transparent=TRANSPARENT,
editable=EDITABLE,
bars=True,
lines=False,
alert=alert,
).auto_ref_ids()
def generate_desired_count_graph(
name: str,
cluster_name: str,
max: int,
cloudwatch_data_source: str,
notifications: List[str],
grid_pos: GridPos,
):
targets = [
CloudwatchMetricsTarget(
alias="Containers",
namespace=CONTAINER_INSIGHTS_NAMESPACE,
statistics=["Maximum"],
metricName="DesiredTaskCount",
dimensions={
"ServiceName": name,
"ClusterName": cluster_name
},
refId=ALERT_REF_ID,
),
]
alert = None
if notifications and max > 1:
alert = Alert(
name="{} Desired count of containers nearing the max".format(name),
message="{} is having Desired count of containers nearing the max".
format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("15m", "now"),
evaluator=GreaterThan(0.9 * max), # 90% of max
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="1m",
notifications=notifications,
)
return Graph(
title="Desired Tasks",
dataSource=cloudwatch_data_source,
targets=targets,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
gridPos=grid_pos,
alertThreshold=ALERT_THRESHOLD,
).auto_ref_ids()
def generate_rds_transaction_id_graph(name: str, cloudwatch_data_source: str,
notifications: List[str]):
"""
Generate rds graph
"""
y_axes = single_y_axis(format=SHORT_FORMAT)
targets = [
CloudwatchMetricsTarget(
alias="Transaction ids used",
metricName="MaximumUsedTransactionIDs",
statistics=["Maximum"],
namespace=NAMESPACE,
dimensions={"DBInstanceIdentifier": name},
period="1m",
refId=ALERT_REF_ID,
),
]
alert = None
if notifications:
alert = Alert(
name="{} transaction ids used Errors".format(name),
message="{} is having transaction ids used errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(1000000000),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="2m",
frequency="2m",
notifications=notifications,
)
return Graph(
title="Transaction ids used",
dataSource=cloudwatch_data_source,
targets=targets,
yAxes=y_axes,
transparent=TRANSPARENT,
editable=EDITABLE,
bars=False,
lines=True,
alert=alert,
).auto_ref_ids()
def create_lambda_sqs_dlq_graph(name: str, cloudwatch_data_source: str,
fifo: bool, notifications: List[str]):
"""Create SQS Deadletter graph"""
if fifo:
name += ".fifo"
targets = [
CloudwatchMetricsTarget(
alias="Approximate number of messages available",
namespace="AWS/SQS",
statistics=["Maximum"],
metricName="ApproximateNumberOfMessagesVisible",
dimensions={"QueueName": name},
refId=ALERT_REF_ID if notifications else None,
)
]
yAxes = single_y_axis(format=SHORT_FORMAT)
alert = None
# https://docs.aws.amazon.com/AWSSimpleQueueService/latest/SQSDeveloperGuide/sqs-monitoring-using-cloudwatch.html
# https://aws.amazon.com/about-aws/whats-new/2019/12/amazon-sqs-now-supports-1-minute-cloudwatch-metrics/
if notifications:
alert = Alert(
name="{} messages".format(name),
message="{} is having messages".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_AND,
),
],
gracePeriod="5m",
notifications=notifications,
)
return Graph(
title="SQS Dead Letter Queue: {}".format(name),
dataSource=cloudwatch_data_source,
targets=targets,
yAxes=yAxes,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
alertThreshold=ALERT_THRESHOLD,
).auto_ref_ids()
def generate_api_gateway_requests_graph(name: str, cloudwatch_data_source: str,
notifications: List[str], *args,
**kwargs):
targets = [
CloudwatchMetricsTarget(
alias=API_GATEWAY_5XX_ALIAS,
namespace=NAMESPACE,
statistics=["Sum"],
metricName="5XXError",
dimensions={"ApiName": name},
refId=ALERT_REF_ID,
),
CloudwatchMetricsTarget(
alias=API_GATEWAY_REQUESTS_ALIAS,
namespace=NAMESPACE,
statistics=["Sum"],
metricName="Count",
dimensions={"ApiName": name},
refId=API_GATEWAY_REQUESTS_REF_ID,
),
CloudwatchMetricsTarget(
alias=API_GATEWAY_4XX_ALIAS,
namespace=NAMESPACE,
statistics=["Sum"],
metricName="4XXError",
dimensions={"ApiName": name},
refId=API_GATEWAY_4XX_REF_ID,
),
]
yAxes = YAxes(
YAxis(format=SHORT_FORMAT),
YAxis(format=SHORT_FORMAT),
)
seriesOverrides = [
{
"alias": API_GATEWAY_REQUESTS_ALIAS,
"points": False,
"color": colors.GREEN,
},
{
"alias": API_GATEWAY_4XX_ALIAS,
"color": colors.YELLOW,
},
{
"alias": API_GATEWAY_5XX_ALIAS,
"color": colors.RED,
},
]
alert = None
# https://docs.aws.amazon.com/lambda/latest/dg/monitoring-metrics.html
if notifications:
alert = Alert(
name="{} API Gateway 5XX Errors".format(name),
message="{} is having 5XX errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("15m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
frequency="2m",
gracePeriod="2m",
notifications=notifications,
)
return Graph(
title="API Gateway Requests: {}".format(name),
dataSource=cloudwatch_data_source,
targets=targets,
seriesOverrides=seriesOverrides,
yAxes=yAxes,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
alertThreshold=ALERT_THRESHOLD,
).auto_ref_ids()
def generate_elasticsearch_jvm_memory_pressure_graph(
name: str, client_id: str, cloudwatch_data_source: str,
notifications: List[str]) -> Graph:
"""
Generate Elasticsearch graph
"""
y_axes = single_y_axis(format=PERCENT_FORMAT)
alias = "JVM memory pressure"
targets = [
CloudwatchMetricsTarget(
alias=alias,
namespace=NAMESPACE,
period="1m",
statistics=["Maximum"],
dimensions={
"DomainName": name,
"ClientId": client_id
},
metricName="JVMMemoryPressure",
refId=ALERT_REF_ID,
)
]
alert = None
if notifications:
alert = Alert(
name="Elasticsearch JVM memory pressure alert",
message="Elasticsearch JVM memory pressure alert",
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(80),
reducerType=RTYPE_MAX,
operator=OP_OR,
),
],
frequency="2m",
gracePeriod="2m",
notifications=notifications,
)
series_overrides = [{
"alias": alias,
"color": colors.GREEN,
"lines": True,
"bars": False,
}]
return Graph(
title=alias,
dataSource=cloudwatch_data_source,
targets=targets,
yAxes=y_axes,
seriesOverrides=series_overrides,
transparent=TRANSPARENT,
editable=EDITABLE,
bars=True,
lines=False,
alert=alert,
).auto_ref_ids()
def generate_res_count_graph(
name: str,
cloudwatch_data_source: str,
loadbalancer: str,
target_group: str,
grid_pos: GridPos,
notifications: List[str],
) -> Graph:
"""
Generate res graph
"""
xx2_alias = "2xx"
xx3_alias = "3xx"
xx4_alias = "4xx"
xx5_alias = "5xx"
targets = [
CloudwatchMetricsTarget(
alias=xx2_alias,
namespace="AWS/ApplicationELB",
statistics=["Sum"],
metricName="HTTPCode_Target_2XX_Count",
dimensions={
"LoadBalancer": loadbalancer,
"TargetGroup": target_group
},
),
CloudwatchMetricsTarget(
alias=xx3_alias,
namespace="AWS/ApplicationELB",
statistics=["Sum"],
metricName="HTTPCode_Target_3XX_Count",
dimensions={
"LoadBalancer": loadbalancer,
"TargetGroup": target_group
},
),
CloudwatchMetricsTarget(
alias=xx4_alias,
namespace="AWS/ApplicationELB",
statistics=["Sum"],
metricName="HTTPCode_Target_4XX_Count",
dimensions={
"LoadBalancer": loadbalancer,
"TargetGroup": target_group
},
),
CloudwatchMetricsTarget(
alias=xx5_alias,
namespace="AWS/ApplicationELB",
statistics=["Sum"],
metricName="HTTPCode_Target_5XX_Count",
dimensions={
"LoadBalancer": loadbalancer,
"TargetGroup": target_group
},
refId=ALERT_REF_ID,
),
]
seriesOverrides = [
{
"alias": xx2_alias,
"color": colors.GREEN,
"fill": 0
},
{
"alias": xx3_alias,
"color": colors.YELLOW,
"fill": 0
},
{
"alias": xx4_alias,
"color": colors.ORANGE,
"fill": 0
},
{
"alias": xx5_alias,
"color": colors.RED,
"fill": 0
},
]
alert = None
if notifications:
alert = Alert(
name="{} has 5XX errors".format(name),
message="{} has 5XX errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("15m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="1m",
notifications=notifications,
)
return Graph(
title="Responses",
dataSource=cloudwatch_data_source,
targets=targets,
seriesOverrides=seriesOverrides,
transparent=TRANSPARENT,
editable=EDITABLE,
gridPos=grid_pos,
alert=alert,
)
def generate_mem_utilization_percentage_graph(
name: str,
cloudwatch_data_source: str,
cluster_name: str,
notifications: List[str],
grid_pos: GridPos,
) -> Graph:
"""
Generate Mem Percentage graph
"""
y_axes = single_y_axis(format=PERCENT_FORMAT)
targets = [
CloudwatchMetricsTarget(
alias=MINIMUM_ALIAS,
namespace=ECS_NAMESPACE,
statistics=["Minimum"],
metricName="MemoryUtilization",
dimensions={
"ServiceName": name,
"ClusterName": cluster_name
},
),
CloudwatchMetricsTarget(
alias=AVERAGE_ALIAS,
namespace=ECS_NAMESPACE,
statistics=["Average"],
metricName="MemoryUtilization",
dimensions={
"ServiceName": name,
"ClusterName": cluster_name
},
refId=ALERT_REF_ID,
),
CloudwatchMetricsTarget(
alias=MAXIMUM_ALIAS,
namespace=ECS_NAMESPACE,
statistics=["Maximum"],
metricName="MemoryUtilization",
dimensions={
"ServiceName": name,
"ClusterName": cluster_name
},
),
]
seriesOverrides = [
{
"alias": MINIMUM_ALIAS,
"color": colors.GREEN,
"lines": False
},
{
"alias": AVERAGE_ALIAS,
"color": colors.YELLOW,
"fill": 0
},
{
"alias": MAXIMUM_ALIAS,
"color": colors.GREEN,
"fillBelowTo": MINIMUM_ALIAS,
"lines": False,
},
]
alert = None
if notifications:
alert = Alert(
name="{} Memory utilization Errors".format(name),
message="{} is having Memory utilization errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("15m", "now"),
evaluator=GreaterThan(85),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="1m",
notifications=notifications,
)
return Graph(
title="Memory Utilization Percentage",
dataSource=cloudwatch_data_source,
targets=targets,
yAxes=y_axes,
seriesOverrides=seriesOverrides,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
gridPos=grid_pos,
alertThreshold=ALERT_THRESHOLD,
).auto_ref_ids()
def generate_rds_cpu_graph(name: str, cloudwatch_data_source: str,
notifications: List[str]):
"""
Generate rds graph
"""
y_axes = single_y_axis(format=PERCENT_FORMAT)
min_alias = "min"
max_alias = "max"
mean_alias = "mean"
targets = [
CloudwatchMetricsTarget(
alias=max_alias,
namespace=NAMESPACE,
dimensions={"DBInstanceIdentifier": name},
period="1m",
statistics=["Maximum"],
metricName="CPUUtilization",
refId=ALERT_REF_ID,
),
CloudwatchMetricsTarget(
alias=mean_alias,
namespace=NAMESPACE,
dimensions={"DBInstanceIdentifier": name},
statistics=["Average"],
metricName="CPUUtilization",
period="1m",
),
CloudwatchMetricsTarget(
alias=min_alias,
namespace=NAMESPACE,
dimensions={"DBInstanceIdentifier": name},
statistics=["Minimum"],
metricName="CPUUtilization",
period="1m",
),
]
series_overrides = get_series_overrides(min_alias, mean_alias, max_alias)
alert = None
if notifications:
alert = Alert(
name="{} CPU utilization Errors".format(name),
message="{} is having CPU utilization errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(80),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="2m",
frequency="2m",
notifications=notifications,
)
return Graph(
title="CPU utilization",
dataSource=cloudwatch_data_source,
targets=targets,
yAxes=y_axes,
seriesOverrides=series_overrides,
transparent=TRANSPARENT,
editable=EDITABLE,
bars=False,
lines=True,
alert=alert,
).auto_ref_ids()
def lambda_generate_memory_utilization_percentage_graph(
name: str,
cloudwatch_data_source: str,
lambda_insights_namespace: str,
notifications: List[str],
*args,
**kwargs,
) -> Graph:
"""
Generate lambda graph
"""
targets = [
CloudwatchMetricsTarget(
alias=MINIMUM_ALIAS,
namespace=lambda_insights_namespace,
statistics=["Minimum"],
metricName="memory_utilization",
dimensions={"function_name": name},
),
CloudwatchMetricsTarget(
alias=AVERAGE_ALIAS,
namespace=lambda_insights_namespace,
statistics=["Average"],
metricName="memory_utilization",
dimensions={"function_name": name},
refId=ALERT_REF_ID,
),
CloudwatchMetricsTarget(
alias=MAXIMUM_ALIAS,
namespace=lambda_insights_namespace,
statistics=["Maximum"],
metricName="memory_utilization",
dimensions={"function_name": name},
),
]
yAxes = YAxes(
YAxis(format=SHORT_FORMAT, decimals=2),
YAxis(format=SHORT_FORMAT, decimals=2),
)
seriesOverrides = [
{
"alias": MINIMUM_ALIAS,
"color": "#C8F2C2",
"lines": False
},
{
"alias": AVERAGE_ALIAS,
"color": "#FADE2A",
"fill": 0
},
{
"alias": MAXIMUM_ALIAS,
"color": "rgb(77, 159, 179)",
"fillBelowTo": MINIMUM_ALIAS,
"lines": False,
},
]
alert = None
# https://docs.aws.amazon.com/lambda/latest/dg/monitoring-metrics.html
if notifications:
alert = Alert(
name="{} Memory utilization Errors".format(name),
message="{} is having Memory utilization errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(90),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="1m",
notifications=notifications,
)
return Graph(
title="Lambda Memory Utilization Percentage",
dataSource=cloudwatch_data_source,
targets=targets,
seriesOverrides=seriesOverrides,
yAxes=yAxes,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
alertThreshold=ALERT_THRESHOLD,
# gridPos=GridPos(8,12,0,0)
).auto_ref_ids()
def lambda_generate_invocations_graph(name: str, cloudwatch_data_source: str,
notifications: List[str], *args,
**kwargs) -> Graph:
"""
Generate lambda graph
"""
targets = [
CloudwatchMetricsTarget(
alias=LAMBDA_INVOCATIONS_ALIAS,
namespace=NAMESPACE,
statistics=["Sum"],
metricName="Invocations",
dimensions={"FunctionName": name},
),
CloudwatchMetricsTarget(
alias=LAMBDA_ERRORS_ALIAS,
namespace=NAMESPACE,
statistics=["Sum"],
metricName="Errors",
dimensions={"FunctionName": name},
refId=ALERT_REF_ID,
),
]
yAxes = YAxes(
YAxis(format=SHORT_FORMAT, decimals=2),
YAxis(format=SHORT_FORMAT, decimals=2),
)
seriesOverrides = [
{
"alias": LAMBDA_INVOCATIONS_ALIAS,
"points": False,
"color": colors.GREEN,
},
{
"alias": LAMBDA_ERRORS_ALIAS,
"points": False,
"color": colors.RED,
},
]
alert = None
# https://docs.aws.amazon.com/lambda/latest/dg/monitoring-metrics.html
if notifications:
alert = Alert(
name="{} Invocation Errors".format(name),
message="{} is having invocation errors".format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=ALERT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_AND,
)
],
gracePeriod="1m",
notifications=notifications,
)
return Graph(
title="Lambda Invocations and Errors",
dataSource=cloudwatch_data_source,
targets=targets,
seriesOverrides=seriesOverrides,
yAxes=yAxes,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
alertThreshold=ALERT_THRESHOLD,
# gridPos=GridPos(8,12,0,0)
).auto_ref_ids()
def generate_sfn_execution_metrics_graph(name: str,
cloudwatch_data_source: str,
notifications: List[str], *args,
**kwargs):
"""
Generate step function graph
"""
targets = [
CloudwatchMetricsTarget(
alias=SFN_EXECUTIONS_STARTED_ALIAS,
namespace=NAMESPACE,
metricName="ExecutionsStarted",
statistics=["Sum"],
dimensions={"StateMachineArn": name},
),
CloudwatchMetricsTarget(
alias=SFN_EXECUTIONS_SUCCEEDED_ALIAS,
namespace=NAMESPACE,
metricName="ExecutionsSucceeded",
statistics=["Sum"],
dimensions={"StateMachineArn": name},
),
CloudwatchMetricsTarget(
alias=SFN_EXECUTIONS_ABORTED_ALIAS,
namespace=NAMESPACE,
metricName="ExecutionsAborted",
statistics=["Sum"],
dimensions={"StateMachineArn": name},
refId=SFN_EXECUTIONS_ABORTED_REF_ID,
),
CloudwatchMetricsTarget(
alias=SFN_EXECUTIONS_FAILED_ALIAS,
namespace=NAMESPACE,
metricName="ExecutionsFailed",
statistics=["Sum"],
dimensions={"StateMachineArn": name},
refId=SFN_EXECUTIONS_FAILED_REF_ID,
),
CloudwatchMetricsTarget(
alias=SFN_EXECUTIONS_THROTTLED_ALIAS,
namespace=NAMESPACE,
metricName="ExecutionsThrottled",
statistics=["Sum"],
dimensions={"StateMachineArn": name},
refId=SFN_EXECUTIONS_THROTTLED_REF_ID,
),
CloudwatchMetricsTarget(
alias=SFN_EXECUTIONS_TIMEDOUT_ALIAS,
namespace=NAMESPACE,
metricName="ExecutionsTimedOut",
statistics=["Sum"],
dimensions={"StateMachineArn": name},
refId=SFN_EXECUTIONS_TIMEDOUT_REF_ID,
),
]
yAxes = YAxes(
YAxis(format=SHORT_FORMAT, decimals=2),
YAxis(format=SHORT_FORMAT, decimals=2),
)
seriesOverrides = [
{
"alias": SFN_EXECUTIONS_STARTED_ALIAS,
"points": False,
"color": colors.BLUE,
},
{
"alias": SFN_EXECUTIONS_SUCCEEDED_ALIAS,
"points": False,
"color": colors.GREEN,
},
{
"alias": SFN_EXECUTIONS_ABORTED_ALIAS,
"points": False,
"color": colors.RED,
},
{
"alias": SFN_EXECUTIONS_FAILED_ALIAS,
"points": False,
"color": colors.RED,
},
{
"alias": SFN_EXECUTIONS_THROTTLED_ALIAS,
"points": False,
"color": colors.ORANGE,
},
{
"alias": SFN_EXECUTIONS_TIMEDOUT_ALIAS,
"points": False,
"color": colors.RED,
},
]
alert = None
# https://docs.aws.amazon.com/lambda/latest/dg/monitoring-metrics.html
if notifications:
alert = Alert(
name="{} execution issues".format(name),
message="{} might have failed, aborted, throttled or timedout".
format(name),
executionErrorState="alerting",
alertConditions=[
AlertCondition(
Target(refId=SFN_EXECUTIONS_ABORTED_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_OR,
),
AlertCondition(
Target(refId=SFN_EXECUTIONS_FAILED_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_OR,
),
AlertCondition(
Target(refId=SFN_EXECUTIONS_THROTTLED_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_OR,
),
AlertCondition(
Target(refId=SFN_EXECUTIONS_TIMEDOUT_REF_ID),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(0),
reducerType=RTYPE_MAX,
operator=OP_OR,
),
],
gracePeriod="2m",
frequency="2m",
notifications=notifications,
)
return Graph(
title="Step function execution metrics",
dataSource=cloudwatch_data_source,
targets=targets,
seriesOverrides=seriesOverrides,
yAxes=yAxes,
transparent=TRANSPARENT,
editable=EDITABLE,
alert=alert,
alertThreshold=ALERT_THRESHOLD,
).auto_ref_ids()
YAxis(format=OPS_FORMAT),
YAxis(format=SHORT_FORMAT),
),
alert=Alert(
name="Too many 500s on Nginx",
message="More than 5 QPS of 500s on Nginx for 5 minutes",
alertConditions=[
AlertCondition(
Target(
expr=
'sum(irate(nginx_http_requests_total{job="default/frontend",status=~"5.."}[1m]))',
legendFormat="5xx",
refId='A',
),
timeRange=TimeRange("5m", "now"),
evaluator=GreaterThan(5),
operator=OP_AND,
reducerType=RTYPE_SUM,
),
],
notifications=[
Notification("notification_channel_uid"),
],
)),
Graph(
title="Frontend latency",
dataSource='My Prometheus',
targets=[
Target(
expr=
'histogram_quantile(0.5, sum(irate(nginx_http_request_duration_seconds_bucket{job="default/frontend"}[1m])) by (le))',