package pulse.config.bootstrap.v1
Get desktop application:
View/edit binary Protocol Buffers messages
message
bootstrap.proto:90Admin server configuration.
Used in:
string bind = 1
The IP/port that the admin server should bind to.
message
bootstrap.proto:21Advanced pipeline configuration.
Used in:
optional uint64 max_cached_metrics = 1
An optional maximum number of metrics (active timeseries) that the proxy will cache. This plays a large part in limiting the maximum memory that the proxy will use and should be scaled to the deployment. It is highly recommended that this be tuned and set. This is a stop gap until the proxy has more dynamic load shed / overload support.
message
bootstrap.proto:158Main bootstrap config for the proxy.
optional admin = 1
Admin configuration.
optional meta_stats = 2
Configuration for output of "meta" stats for the proxy. These are metrics that get sent directly to a target without going through the rest of the pipeline.
optional kubernetes = 3
Bootstrap configuration for Kubernetes integration.
oneof pipeline_type
pipeline = 4
Static pipeline configured in the bootstrap.
fs_watched_pipeline = 5
Pipeline that will be read and watched from the filesystem (e.g., K8s config maps). Reloads are meant to occur on atomic symbolic link swaps.
merged_pipeline = 6
This configuration option combines the previous two options and allows for the most flexibility. It allows for a bootstrap pipeline to be merged with any number of FS watched pipelines. During merging, if a pipeline component with the same name is found later in the merging process it will override anything found previously to form the final pipeline. After merging the final configuration checks will take place to make sure that routes are valid and so on. Technically the previous options can be expressed as this option but they are split out for clarity.
message
bootstrap.proto:96Bootstrap configuration for Kubernetes integration.
Used in:
optional node_name = 1
Supplies the node name used for K8s API lookups against the node.
optional bool evaluate_services = 2
Determines whether service evaluation will occur, typically to look for Prometheus scrape labels on the service. If only pod labels/annotations are being used this can be disabled. Defaults to true.
optional services_cache_interval = 4
Services are fetched by namespace on demand, and then cached, with the assumption that they rarely change. This configures the cache time. The default is 15 minutes.
repeated string pod_phases = 3
Specified which pod phases should be included in the cache, and as a result scraped, used for metadata attachment, etc. Defaults to ["Pending", "Running"] if empty. "Pending" is included by default (as long as there is any attached IP address) so that pulse knows about a pod as soon as possible and it can also begin scraping to provide an "up" metric if desired. This can be used to monitor pods stuck in pending if desired.
optional host_network_pod_by_ip_filter = 5
If associating pods by source IP address (for example when receiving statsd from a pod), behavior is undefined if multiple pods are on the same host network. This filter allows specifying a label or annotation that must be present on the pod in order for it to be associated with the source IP address. This is useful for ensuring that only one pod on a host network is associated with a given source IP address. A warning and metric will be emitted if this is not set and multiple pods are found on the same host network. NOTE: This is ONLY used when the pod is on a host network. It does not apply to other pods.
message
bootstrap.proto:105Used in:
oneof filter_type
label_matcher = 1
Matches against the pod labels.
annotation_matcher = 2
Matches against the pod annotations.
message
bootstrap.proto:97Used in:
string name = 1
The metadata name to match against.
string value_regex = 2
The regex to match the metadata value against.
message
bootstrap.proto:148Configuration composed of layers.
Used in:
optional bootstrap = 1
Optional inline bootstrap configuration.
repeated fs_watched_pipelines = 2
Any number of configuration layers. Later layers in the list will override earlier layers if any names are shared.
message
bootstrap.proto:46Configuration for output of "meta" stats for the proxy. These are metrics that get sent directly to a target without going through the rest of the pipeline.
Used in:
repeated meta_protocol = 1
Which protocol to use for sending meta stats. Multiple protocols can be configured to send data to multiple targets.
optional string meta_prefix = 2
An optional prefix to use on each meta stat name. Defaults to pulse_proxy if not set.
repeated meta_tag = 3
The set of tags to send with each meta stat.
optional node_id = 4
The node ID used to populate the "source" meta stat tag.
optional flush_interval = 5
Flush interval for meta stats. Defaults to 60s. Note: This interval is rounded to the nearest second, and is pegged to *wall clock time.* A setting of 60s will attempt to fire as close to the 1st second of every minute as possible. A setting of 5s will attempt to fire at the [0, 5, 10, ...] seconds of every minute.
message
bootstrap.proto:48Protocol configuration for meta stats export.
Used in:
oneof protocol
wire = 1
Configuration for a TCP based export (statsd or carbon).
prom_remote_write = 2
Configuration for a Prometheus remote write export.
message
bootstrap.proto:61Tags sent with each meta stat.
Used in:
string key = 1
Tag key. Cannot be "source." Set "source" using the node ID.
string value = 2
Tag value.
message
bootstrap.proto:30Configuration for a pipeline.
Used in: ,
map<string, > inflows = 1
All inflow configurations.
map<string, > processors = 2
All processor configurations.
map<string, > outflows = 3
All outflow configurations.
optional advanced = 4
Advanced configuration for the pipeline.