This is release v0.35.0-rc.0 of Grafana Agent.
Upgrading
Read the relevant upgrade guides for specific instructions on upgrading from older versions:
Notable changes:
Breaking changes
-
The algorithm for the "hash" action of
otelcol.processor.attributeshas changed.
The change was made in PR #22831 of opentelemetry-collector-contrib. (@ptodev) -
otelcol.exporter.lokinow includes the instrumentation scope in its output. (@ptodev) -
otelcol.extension.jaeger_remote_samplingremoves the/HTTP endpoint. The/samplingendpoint is still functional.
The change was made in PR #18070 of opentelemetry-collector-contrib. (@ptodev) -
The field
versionandauthstruct block fromwalk_paramsinprometheus.exporter.snmpand SNMP integration have been removed. The auth block now can be configured at top level, together withmodules(@marctc) -
Rename
discovery.filetolocal.file_matchto make it more clear that it
discovers file on the local filesystem, and so it doesn't get confused with
Prometheus' file discovery. (@rfratto) -
Remove the
discovery_target_decodefunction in favor of using discovery
components to better match the behavior of Prometheus' service discovery.
(@rfratto) -
In the traces subsystem for Static mode, some metrics are removed and others are renamed. (@ptodev)
-
Removed metrics:
- "blackbox_exporter_config_last_reload_success_timestamp_seconds" (gauge)
- "blackbox_exporter_config_last_reload_successful" (gauge)
- "blackbox_module_unknown_total" (counter)
- "traces_processor_tail_sampling_count_traces_sampled" (counter)
- "traces_processor_tail_sampling_new_trace_id_received" (counter)
- "traces_processor_tail_sampling_sampling_decision_latency" (histogram)
- "traces_processor_tail_sampling_sampling_decision_timer_latency" (histogram)
- "traces_processor_tail_sampling_sampling_policy_evaluation_error" (counter)
- "traces_processor_tail_sampling_sampling_trace_dropped_too_early" (counter)
- "traces_processor_tail_sampling_sampling_traces_on_memory" (gauge)
- "traces_receiver_accepted_spans" (counter)
- "traces_receiver_refused_spans" (counter)
- "traces_exporter_enqueue_failed_log_records" (counter)
- "traces_exporter_enqueue_failed_metric_points" (counter)
- "traces_exporter_enqueue_failed_spans" (counter)
- "traces_exporter_queue_capacity" (gauge)
- "traces_exporter_queue_size" (gauge)
-
Renamed metrics:
- "traces_receiver_refused_spans" is renamed to "traces_receiver_refused_spans_total"
- "traces_receiver_accepted_spans" is renamed to "traces_receiver_refused_spans_total"
- "traces_exporter_sent_metric_points" is renamed to "traces_exporter_sent_metric_points_total"
-
-
The
remote_samplingblock has been removed fromotelcol.receiver.jaeger. (@ptodev)
Deprecations
otelcol.exporter.jaegerhas been deprecated and will be removed in Agent v0.38.0. (@ptodev)
Features
-
The Pyroscope scrape component computes and sends delta profiles automatically when required to reduce bandwidth usage. (@cyriltovena)
-
Support
stage.geoipinloki.process. (@akselleirv) -
Integrations: Introduce the
squidintegration. (@armstrmi) -
Support custom fields in MMDB file for
stage.geoip. (@akselleirv) -
Added json_path function to river stdlib. (@jkroepke)
-
Add
format,join,tp_lower,replace,split,trim,trim_prefix,trim_suffix,trim_space,to_upperfunctions to river stdlib. (@jkroepke) -
Flow UI: Add a view for listing the Agent's peers status when clustering is enabled. (@tpaschalis)
-
Add a new CLI command
grafana-agent convertfor converting a river file from supported formats to river. (@erikbaranowski) -
Add support to the
grafana-agent runCLI for converting a river file from supported formats to river. (@erikbaranowski) -
Add boringcrypto builds and docker images for Linux arm64 and x64. (@mattdurham)
-
New Grafana Agent Flow components:
discovery.filediscovers scrape targets from files. (@spartan0x117)discovery.kubeletcollect scrape targets from the Kubelet API. (@gcampbell12)module.httpruns a Grafana Agent Flow module loaded from a remote HTTP endpoint. (@spartan0x117)otelcol.processor.attributesaccepts telemetry data from otherotelcol
components and modifies attributes of a span, log, or metric. (@ptodev)prometheus.exporter.cloudwatch- scrape AWS CloudWatch metrics (@thepalbi)prometheus.exporter.elasticsearchcollects metrics from Elasticsearch. (@marctc)prometheus.exporter.kafkacollects metrics from Kafka Server. (@oliver-zhang)prometheus.exporter.mongodbcollects metrics from MongoDB. (@marctc)prometheus.exporter.squidcollects metrics from a squid server. (@armstrmi)prometheus.operator.probes- discovers Probe resources in your Kubernetes
cluster and scrape the targets they reference. (@captncraig)pyroscope.ebpfcollects system-wide performance profiles from the current
host (@korniltsev)otelcol.exporter.loadbalancing- export traces and logs to multiple OTLP gRPC
endpoints in a load-balanced way. (@ptodev)
-
New Grafana Agent Flow command line utilities:
grafana-agent tools prometheus.remote_writeholds a collection of remote
write-specific tools. These have been ported over from theagentctlcommand. (@rfratto)
-
A new
actionargument forotelcol.auth.headers. (@ptodev) -
New
metadata_keysandmetadata_cardinality_limitarguments forotelcol.processor.batch. (@ptodev) -
New
boolean_attributeandottl_conditionsampling policies forotelcol.processor.tail_sampling. (@ptodev) -
A new
initial_offsetargument forotelcol.receiver.kafka. (@ptodev)
For a full list of changes, please refer to the CHANGELOG!
Installation
Refer to our installation guides for how to install the variants of Grafana Agent: