gcp.vertex-ai-index-endpoint
GCP Vertex AI Vector Search Index Endpoint Resource
Index endpoints serve Vector Search indexes. Each deployed index runs dedicated replicas, billed whether or not the index receives queries.
Metrics match per endpoint, so an endpoint serving one busy and one idle index counts as busy.
- example:
Find index endpoints with a deployed index and no queries in 7 days:
policies:
- name: vertexai-index-endpoint-idle
resource: gcp.vertex-ai-index-endpoint
filters:
- type: value
key: deployedIndexes
value: not-null
- type: metrics
name: aiplatform.googleapis.com/matching_engine/query/request_count
aligner: ALIGN_SUM
reducer: REDUCE_SUM
group-by-fields:
- resource.labels.index_endpoint_id
days: 7
op: lte
value: 0
missing-value: 0
Filters
metrics
Supports metrics filters on resources.
All resources that have cloud watch metrics are supported.
Docs on cloud watch metrics
Google Supported Metrics https://cloud.google.com/monitoring/api/metrics_gcp
Custom Metrics https://cloud.google.com/monitoring/api/v3/metric-model#intro-custom-metrics
- name: firewall-hit-count
resource: gcp.firewall
filters:
- type: metrics
name: firewallinsights.googleapis.com/subnet/firewall_hit_count
aligner: ALIGN_COUNT
days: 14
value: 1
op: greater-than
The period-start key allows you to align the metric window in two ways.
By default, using auto, the window is computed relative to the current time.
Alternatively, setting it to start-of-day aligns the window to full UTC calendar days,
beginning at 00:00:00 UTC and ending at current day 00:00:00 UTC.
- name: instance-low-cpu-last-full-day
resource: gcp.instance
filters:
- type: metrics
name: compute.googleapis.com/instance/cpu/utilization
aligner: ALIGN_MEAN
days: 1
value: 0.05
op: less-than
period-start: start-of-day
Cloud Monitoring returns a DISTRIBUTION-typed metric as a histogram rather
than a number. The distribution key chooses how to reduce one to a scalar:
count, mean, or sum (the default), reconstructed as count * mean
since a histogram carries no total. Scalar metrics ignore the key. Either way
the filter compares the most recent point in the window, so a metric left on
the default ALIGN_NONE is read on its latest sample, not its total.
- name: tuned-model-chatty-responses
resource: gcp.vertex-ai-endpoint
filters:
- type: metrics
name: aiplatform.googleapis.com/tuned_model/online_serving/tokens
aligner: ALIGN_SUM
reducer: REDUCE_SUM
group-by-fields: ["resource.labels.endpoint_id"]
distribution: mean
days: 1
value: 100
op: greater-than
properties:
aligner:
enum:
- ALIGN_NONE
- ALIGN_DELTA
- ALIGN_RATE
- ALIGN_INTERPOLATE
- ALIGN_MIN
- ALIGN_MAX
- ALIGN_MEAN
- ALIGN_COUNT
- ALIGN_SUM
- REDUCE_COUNT_FALSE
- ALIGN_STDDEV
- ALIGN_COUNT_TRUE
- ALIGN_COUNT_FALSE
- ALIGN_FRACTION_TRUE
- ALIGN_PERCENTILE_99
- ALIGN_PERCENTILE_95
- ALIGN_PERCENTILE_50
- ALIGN_PERCENTILE_05
- ALIGN_PERCENT_CHANG
type: string
days:
type: number
distribution:
enum:
- count
- mean
- sum
type: string
filter:
type: string
group-by-fields:
items:
type: string
type: array
metric-key:
type: string
missing-value:
type: number
name:
type: string
op:
enum:
- eq
- equal
- ne
- not-equal
- gt
- greater-than
- ge
- gte
- le
- lte
- lt
- less-than
- glob
- regex
- regex-case
- in
- ni
- not-in
- contains
- difference
- intersect
- mod
type: string
period-start:
enum:
- auto
- start-of-day
type: string
reducer:
enum:
- REDUCE_NONE
- REDUCE_MEAN
- REDUCE_MIN
- REDUCE_MAX
- REDUCE_MEAN
- REDUCE_SUM
- REDUCE_STDDEV
- REDUCE_COUNT
- REDUCE_COUNT_TRUE
- REDUCE_COUNT_FALSE
- REDUCE_FRACTION_TRUE
- REDUCE_PERCENTILE_99
- REDUCE_PERCENTILE_95
- REDUCE_PERCENTILE_50
- REDUCE_PERCENTILE_05
type: string
type:
enum:
- metrics
value:
type: number
required:
- value
- name
- op
Permissions - monitoring.timeSeries.list