A Service-Level Objective (SLO) describes the level of desired good service. It consists of a service-level indicator (SLI), a performance goal, and a period over which the objective is to be evaluated against that goal. The SLO can use SLIs defined in a number of different manners. Typical SLOs might include "99% of requests in each rolling week have latency below 200 milliseconds" or "99.5% of requests in each calendar month return successfully."
To get more information about Slo, see:
data "google_monitoring_app_engine_service" "default" {
module_id = "default"
}
resource "google_monitoring_slo" "appeng_slo" {
service = data.google_monitoring_app_engine_service.default.service_id
slo_id = "ae-slo"
display_name = "Terraform Test SLO for App Engine"
goal = 0.9
calendar_period = "DAY"
basic_sli {
latency {
threshold = "1s"
}
}
user_labels = {
my_key = "my_value"
my_other_key = "my_other_value"
}
}
resource "google_monitoring_custom_service" "customsrv" {
service_id = "custom-srv-request-slos"
display_name = "My Custom Service"
}
resource "google_monitoring_slo" "request_based_slo" {
service = google_monitoring_custom_service.customsrv.service_id
slo_id = "consumed-api-slo"
display_name = "Terraform Test SLO with request based SLI (good total ratio)"
goal = 0.9
rolling_period_days = 30
request_based_sli {
distribution_cut {
distribution_filter = "metric.type=\"serviceruntime.googleapis.com/api/request_latencies\" resource.type=\"api\" "
range {
max = 0.5
}
}
}
}
resource "google_monitoring_custom_service" "customsrv" {
service_id = "custom-srv-windows-slos"
display_name = "My Custom Service"
}
resource "google_monitoring_slo" "windows_based" {
service = google_monitoring_custom_service.customsrv.service_id
display_name = "Terraform Test SLO with window based SLI"
goal = 0.95
calendar_period = "FORTNIGHT"
windows_based_sli {
window_period = "400s"
good_bad_metric_filter = join(" AND ", [
"metric.type=\"monitoring.googleapis.com/uptime_check/check_passed\"",
"resource.type=\"uptime_url\"",
])
}
}
resource "google_monitoring_custom_service" "customsrv" {
service_id = "custom-srv-windows-slos"
display_name = "My Custom Service"
}
resource "google_monitoring_slo" "windows_based" {
service = google_monitoring_custom_service.customsrv.service_id
display_name = "Terraform Test SLO with window based SLI"
goal = 0.9
rolling_period_days = 20
windows_based_sli {
window_period = "600s"
metric_mean_in_range {
time_series = join(" AND ", [
"metric.type=\"agent.googleapis.com/cassandra/client_request/latency/95p\"",
"resource.type=\"gce_instance\"",
])
range {
max = 5
}
}
}
}
resource "google_monitoring_custom_service" "customsrv" {
service_id = "custom-srv-windows-slos"
display_name = "My Custom Service"
}
resource "google_monitoring_slo" "windows_based" {
service = google_monitoring_custom_service.customsrv.service_id
display_name = "Terraform Test SLO with window based SLI"
goal = 0.9
rolling_period_days = 20
windows_based_sli {
window_period = "400s"
metric_sum_in_range {
time_series = join(" AND ", [
"metric.type=\"monitoring.googleapis.com/uptime_check/request_latency\"",
"resource.type=\"uptime_url\"",
])
range {
max = 5000
}
}
}
}
resource "google_monitoring_custom_service" "customsrv" {
service_id = "custom-srv-windows-slos"
display_name = "My Custom Service"
}
resource "google_monitoring_slo" "windows_based" {
service = google_monitoring_custom_service.customsrv.service_id
display_name = "Terraform Test SLO with window based SLI"
goal = 0.9
rolling_period_days = 20
windows_based_sli {
window_period = "100s"
good_total_ratio_threshold {
threshold = 0.1
performance {
distribution_cut {
distribution_filter = join(" AND ", [
"metric.type=\"serviceruntime.googleapis.com/api/request_latencies\"",
"resource.type=\"consumed_api\"",
])
range {
min = 1
max = 9
}
}
}
}
}
}
The following arguments are supported:
goal
-
(Required)
The fraction of service that must be good in order for this objective
to be met. 0 < goal <= 0.999
service
-
(Required)
ID of the service to which this SLO belongs.
display_name
-
(Optional)
Name used for UI elements listing this SLO.
rolling_period_days
-
(Optional)
A rolling time period, semantically "in the past X days".
Must be between 1 to 30 days, inclusive.
calendar_period
-
(Optional)
A calendar period, semantically "since the start of the current
DAY
, WEEK
, FORTNIGHT
, MONTH
.
user_labels
-
(Optional)
This field is intended to be used for organizing and identifying the AlertPolicy
objects.The field can contain up to 64 entries. Each key and value is limited
to 63 Unicode characters or 128 bytes, whichever is smaller. Labels and values
can contain only lowercase letters, numerals, underscores, and dashes. Keys
must begin with a letter.
basic_sli
-
(Optional)
Basic Service-Level Indicator (SLI) on a well-known service type.
Performance will be computed on the basis of pre-defined metrics.
SLIs are used to measure and calculate the quality of the Service's
performance with respect to a single aspect of service quality.
Exactly one of the following must be set:
basic_sli
, request_based_sli
, windows_based_sli
Structure is documented below.
request_based_sli
-
(Optional)
A request-based SLI defines a SLI for which atomic units of
service are counted directly.
A SLI describes a good service.
It is used to measure and calculate the quality of the Service's
performance with respect to a single aspect of service quality.
Exactly one of the following must be set:
basic_sli
, request_based_sli
, windows_based_sli
Structure is documented below.
windows_based_sli
-
(Optional)
A windows-based SLI defines the criteria for time windows.
good_service is defined based off the count of these time windows
for which the provided service was of good quality.
A SLI describes a good service. It is used to measure and calculate
the quality of the Service's performance with respect to a single
aspect of service quality.
Exactly one of the following must be set:
basic_sli
, request_based_sli
, windows_based_sli
Structure is documented below.
slo_id
-
(Optional)
The id to use for this ServiceLevelObjective. If omitted, an id will be generated instead.
project
- (Optional) The ID of the project in which the resource belongs.
If it is not provided, the provider project is used.
method
-
(Optional)
An optional set of RPCs to which this SLI is relevant.
Telemetry from other methods will not be used to calculate
performance for this SLI. If omitted, this SLI applies to all
the Service's methods. For service types that don't support
breaking down by method, setting this field will result in an
error.
location
-
(Optional)
An optional set of locations to which this SLI is relevant.
Telemetry from other locations will not be used to calculate
performance for this SLI. If omitted, this SLI applies to all
locations in which the Service has activity. For service types
that don't support breaking down by location, setting this
field will result in an error.
version
-
(Optional)
The set of API versions to which this SLI is relevant.
Telemetry from other API versions will not be used to
calculate performance for this SLI. If omitted,
this SLI applies to all API versions. For service types
that don't support breaking down by version, setting this
field will result in an error.
latency
-
(Optional)
Parameters for a latency threshold SLI.
Structure is documented below.
availability
-
(Optional)
Availability based SLI, dervied from count of requests made to this service that return successfully.
Structure is documented below.
threshold
-
(Required)
A duration string, e.g. 10s.
Good service is defined to be the count of requests made to
this service that return in no more than threshold.The availability
block supports:
enabled
-
(Optional)
Whether an availability SLI is enabled or not. Must be set to true. Defaults to true
.The request_based_sli
block supports:
good_total_ratio
-
(Optional)
A means to compute a ratio of good_service
to total_service
.
Defines computing this ratio with two TimeSeries monitoring filters
Must specify exactly two of good, bad, and total service filters.
The relationship good_service + bad_service = total_service
will be assumed.
Exactly one of distribution_cut
or good_total_ratio
can be set.
Structure is documented below.
distribution_cut
-
(Optional)
Used when good_service is defined by a count of values aggregated in a
Distribution that fall into a good range. The total_service is the
total count of all values aggregated in the Distribution.
Defines a distribution TimeSeries filter and thresholds used for
measuring good service and total service.
Exactly one of distribution_cut
or good_total_ratio
can be set.
Structure is documented below.
The good_total_ratio
block supports:
good_service_filter
-
(Optional)
A TimeSeries monitoring filter
quantifying good service provided.
Must have ValueType = DOUBLE or ValueType = INT64 and
must have MetricKind = DELTA or MetricKind = CUMULATIVE.
Exactly two of good_service_filter
,bad_service_filter
,total_service_filter
must be set (good + bad = total is assumed).
bad_service_filter
-
(Optional)
A TimeSeries monitoring filter
quantifying bad service provided, either demanded service that
was not provided or demanded service that was of inadequate
quality.
Must have ValueType = DOUBLE or ValueType = INT64 and
must have MetricKind = DELTA or MetricKind = CUMULATIVE.
Exactly two of good_service_filter
,bad_service_filter
,total_service_filter
must be set (good + bad = total is assumed).
total_service_filter
-
(Optional)
A TimeSeries monitoring filter
quantifying total demanded service.
Must have ValueType = DOUBLE or ValueType = INT64 and
must have MetricKind = DELTA or MetricKind = CUMULATIVE.
Exactly two of good_service_filter
,bad_service_filter
,total_service_filter
must be set (good + bad = total is assumed).
The distribution_cut
block supports:
distribution_filter
-
(Required)
A TimeSeries monitoring filter
aggregating values to quantify the good service provided.
Must have ValueType = DISTRIBUTION and
MetricKind = DELTA or MetricKind = CUMULATIVE.
range
-
(Required)
Range of numerical values. The computed good_service
will be the count of values x in the Distribution such
that range.min <= x <= range.max. inclusive of min and
max. Open ranges can be defined by setting
just one of min or max.
Structure is documented below.
min
-
(Optional)
Min value for the range (inclusive). If not given,
will be set to 0
max
-
(Optional)
max value for the range (inclusive). If not given,
will be set to 0
The windows_based_sli
block supports:
window_period
-
(Optional)
Duration over which window quality is evaluated, given as a
duration string "{X}s" representing X seconds. Must be an
integer fraction of a day and at least 60s.
good_bad_metric_filter
-
(Optional)
A TimeSeries monitoring filter
with ValueType = BOOL. The window is good if any true values
appear in the window. One of good_bad_metric_filter
,
good_total_ratio_threshold
, metric_mean_in_range
,
metric_sum_in_range
must be set for windows_based_sli
.
good_total_ratio_threshold
-
(Optional)
Criterion that describes a window as good if its performance is
high enough. One of good_bad_metric_filter
,
good_total_ratio_threshold
, metric_mean_in_range
,
metric_sum_in_range
must be set for windows_based_sli
.
Structure is documented below.
metric_mean_in_range
-
(Optional)
Criterion that describes a window as good if the metric's value
is in a good range, averaged across returned streams.
One of good_bad_metric_filter
,
good_total_ratio_threshold
, metric_mean_in_range
,
metric_sum_in_range
must be set for windows_based_sli
.
Average value X of time_series
should satisfy
range.min <= X <= range.max
for a good window.
Structure is documented below.
metric_sum_in_range
-
(Optional)
Criterion that describes a window as good if the metric's value
is in a good range, summed across returned streams.
Summed value X
of time_series
should satisfy
range.min <= X <= range.max
for a good window.
One of good_bad_metric_filter
,
good_total_ratio_threshold
, metric_mean_in_range
,
metric_sum_in_range
must be set for windows_based_sli
.
Structure is documented below.
The good_total_ratio_threshold
block supports:
threshold
-
(Optional)
If window performance >= threshold, the window is counted
as good.
performance
-
(Optional)
Request-based SLI to evaluate to judge window quality.
Structure is documented below.
basic_sli_performance
-
(Optional)
Basic SLI to evaluate to judge window quality.
Structure is documented below.
The performance
block supports:
good_total_ratio
-
(Optional)
A means to compute a ratio of good_service
to total_service
.
Defines computing this ratio with two TimeSeries monitoring filters
Must specify exactly two of good, bad, and total service filters.
The relationship good_service + bad_service = total_service
will be assumed.
Structure is documented below.
distribution_cut
-
(Optional)
Used when good_service is defined by a count of values aggregated in a
Distribution that fall into a good range. The total_service is the
total count of all values aggregated in the Distribution.
Defines a distribution TimeSeries filter and thresholds used for
measuring good service and total service.
Structure is documented below.
The good_total_ratio
block supports:
good_service_filter
-
(Optional)
A TimeSeries monitoring filter
quantifying good service provided. Exactly two of
good, bad, or total service filter must be defined (where
good + bad = total is assumed)
Must have ValueType = DOUBLE or ValueType = INT64 and
must have MetricKind = DELTA or MetricKind = CUMULATIVE.
bad_service_filter
-
(Optional)
A TimeSeries monitoring filter
quantifying bad service provided, either demanded service that
was not provided or demanded service that was of inadequate
quality. Exactly two of
good, bad, or total service filter must be defined (where
good + bad = total is assumed)
Must have ValueType = DOUBLE or ValueType = INT64 and
must have MetricKind = DELTA or MetricKind = CUMULATIVE.
total_service_filter
-
(Optional)
A TimeSeries monitoring filter
quantifying total demanded service. Exactly two of
good, bad, or total service filter must be defined (where
good + bad = total is assumed)
Must have ValueType = DOUBLE or ValueType = INT64 and
must have MetricKind = DELTA or MetricKind = CUMULATIVE.
The distribution_cut
block supports:
distribution_filter
-
(Required)
A TimeSeries monitoring filter
aggregating values to quantify the good service provided.
Must have ValueType = DISTRIBUTION and
MetricKind = DELTA or MetricKind = CUMULATIVE.
range
-
(Required)
Range of numerical values. The computed good_service
will be the count of values x in the Distribution such
that range.min <= x <= range.max. inclusive of min and
max. Open ranges can be defined by setting
just one of min or max.
Structure is documented below.
min
-
(Optional)
Min value for the range (inclusive). If not given,
will be set to 0
max
-
(Optional)
max value for the range (inclusive). If not given,
will be set to 0
The basic_sli_performance
block supports:
method
-
(Optional)
An optional set of RPCs to which this SLI is relevant.
Telemetry from other methods will not be used to calculate
performance for this SLI. If omitted, this SLI applies to all
the Service's methods. For service types that don't support
breaking down by method, setting this field will result in an
error.
location
-
(Optional)
An optional set of locations to which this SLI is relevant.
Telemetry from other locations will not be used to calculate
performance for this SLI. If omitted, this SLI applies to all
locations in which the Service has activity. For service types
that don't support breaking down by location, setting this
field will result in an error.
version
-
(Optional)
The set of API versions to which this SLI is relevant.
Telemetry from other API versions will not be used to
calculate performance for this SLI. If omitted,
this SLI applies to all API versions. For service types
that don't support breaking down by version, setting this
field will result in an error.
latency
-
(Optional)
Parameters for a latency threshold SLI.
Structure is documented below.
availability
-
(Optional)
Availability based SLI, dervied from count of requests made to this service that return successfully.
Structure is documented below.
threshold
-
(Required)
A duration string, e.g. 10s.
Good service is defined to be the count of requests made to
this service that return in no more than threshold.The availability
block supports:
enabled
-
(Optional)
Whether an availability SLI is enabled or not. Must be set to true. Defaults to
true`.The metric_mean_in_range
block supports:
time_series
-
(Required)
A monitoring filter
specifying the TimeSeries to use for evaluating window
The provided TimeSeries must have ValueType = INT64 or
ValueType = DOUBLE and MetricKind = GAUGE. Mean value X
should satisfy range.min <= X <= range.max
under good service.
range
-
(Required)
Range of numerical values. The computed good_service
will be the count of values x in the Distribution such
that range.min <= x <= range.max. inclusive of min and
max. Open ranges can be defined by setting
just one of min or max. Mean value X
of time_series
values should satisfy range.min <= X <= range.max
for a
good service.
Structure is documented below.
min
-
(Optional)
Min value for the range (inclusive). If not given,
will be set to "-infinity", defining an open range
"< range.max"
max
-
(Optional)
max value for the range (inclusive). If not given,
will be set to "infinity", defining an open range
">= range.min"
The metric_sum_in_range
block supports:
time_series
-
(Required)
A monitoring filter
specifying the TimeSeries to use for evaluating window
quality. The provided TimeSeries must have
ValueType = INT64 or ValueType = DOUBLE and
MetricKind = GAUGE.
Summed value X
should satisfy
range.min <= X <= range.max
for a good window.
range
-
(Required)
Range of numerical values. The computed good_service
will be the count of values x in the Distribution such
that range.min <= x <= range.max. inclusive of min and
max. Open ranges can be defined by setting
just one of min or max. Summed value X
should satisfy
range.min <= X <= range.max
for a good window.
Structure is documented below.
min
-
(Optional)
Min value for the range (inclusive). If not given,
will be set to "-infinity", defining an open range
"< range.max"
max
-
(Optional)
max value for the range (inclusive). If not given,
will be set to "infinity", defining an open range
">= range.min"
In addition to the arguments listed above, the following computed attributes are exported:
id
- an identifier for the resource with format {{name}}
name
-
The full resource name for this service. The syntax is:
projects/[PROJECT_ID_OR_NUMBER]/services/[SERVICE_ID]/serviceLevelObjectives/[SLO_NAME]
This resource provides the following Timeouts configuration options:
create
- Default is 20 minutes.update
- Default is 20 minutes.delete
- Default is 20 minutes.Slo can be imported using any of these accepted formats:
{{name}}
In Terraform v1.5.0 and later, use an import
block to import Slo using one of the formats above. For example:
import {
id = "{{name}}"
to = google_monitoring_slo.default
}
When using the terraform import
command, Slo can be imported using one of the formats above. For example:
$ terraform import google_monitoring_slo.default {{name}}
This resource supports User Project Overrides.