Skip to content

Latest commit

 

History

History
2020 lines (1319 loc) · 115 KB

CHANGELOG.md

File metadata and controls

2020 lines (1319 loc) · 115 KB

Changelog

Contributors should read our contributors guide for instructions on how to update the changelog.

This document contains a historical list of changes between releases. Only changes that impact end-user behavior are listed; changes to documentation or internal API changes are not present.

Main (unreleased)

Enhancements

  • update promtail converter to use file_match block for loki.source.file instead of going through local.file_match. (@kalleep)

v1.12.0-rc.0

Breaking changes

  • prometheus.exporter.blackbox, prometheus.exporter.snmp and prometheus.exporter.statsd now use the component ID instead of the hostname as their instance label in their exported metrics. This is a consequence of a bug fix that could lead to missing data when using the exporter with clustering. If you would like to retain the previous behaviour, you can use discovery.relabel with action = "replace" rule to set the instance label to sys.env("HOSTNAME"). (@thampiotr)

Features

  • (Experimental) Add an otelcol.receiver.cloudflare component to receive logs pushed by Cloudflare's LogPush jobs. (@x1unix)

  • (Experimental) Additions to experimental database_observability.mysql component:

    • explain_plans
      • collector now changes schema before returning the connection to the pool (@cristiangreco)
      • collector now passes queries more permissively, expressly to allow queries beginning in with (@rgeyer)
    • enable explain_plans collector by default (@rgeyer)

  • (Experimental) Additions to experimental database_observability.postgres component:

    • explain_plans
      • added the explain plan collector (@rgeyer)
      • collector now passes queries more permissively, expressly to allow queries beginning in with (@rgeyer)
    • query_samples
      • add user field to wait events within query_samples collector (@gaantunes)
      • rework the query samples collector to buffer per-query execution state across scrapes and emit finalized entries (@gaantunes)
      • process turned idle rows to calculate finalization times precisely and emit first seen idle rows (@gaantunes)
    • query_details
      • escape queries coming from pg_stat_statements with quotes (@gaantunes)
    • enable explain_plans collector by default (@rgeyer)
    • safely generate server_id when UDP socket used for database connection (@matthewnolf)
    • add table registry and include "validated" in parsed table name logs (@fridgepoet)

  • Add otelcol.exporter.googlecloudpubsub community component to export metrics, traces, and logs to Google Cloud Pub/Sub topic. (@eraac)

  • Add structured_metadata_drop stage for loki.process to filter structured metadata. (@baurmatt)

  • Send remote config status to the remote server for the remotecfg service. (@erikbaranowski)

  • Send effective config to the remote server for the remotecfg service. (@erikbaranowski)

  • Add a stat_statements configuration block to the prometheus.exporter.postgres component to enable selecting both the query ID and the full SQL statement. The new block includes one option to enable statement selection, and another to configure the maximum length of the statement text. (@SimonSerrano)

  • Add truncate stage for loki.process to truncate log entries, label values, and structured_metadata values. (@dehaansa)

  • Add u_probe_links & load_probe configuration fields to alloy pyroscope.ebpf to extend configuration of the opentelemetry-ebpf-profiler to allow uprobe profiling and dynamic probing. (@luweglarz)

  • Add verbose_mode configuration fields to alloy pyroscope.ebpf to be enable ebpf-profiler verbose mode. (@luweglarz)

  • Add file_match block to loki.source.file for built-in file discovery using glob patterns. (@kalleep)

  • Add a regex argument to the structured_metadata stage in loki.process to extract labels matching a regular expression. (@timonegk)

  • OpenTelemetry Collector dependencies upgraded from v0.134.0 to v0.139.0. (@dehaansa)

    • All otelcol.receiver.* components leveraging an HTTP server can configure HTTP keep alive behavior with keep_alives_enabled.
    • All otelcol.exporter.* components providing the sending_queue > batch block have default batch values.
    • The otelcol.processor.k8sattributes component has support for extracting annotations from k8s jobs and daemonsets.
    • The otelcol.processor.resourcedecetion component supports nine new detectors.
    • The otelcol.exporter.kafka component supports partitioning logs by trace ID (partition_logs_by_trace_id) and configuring default behavior if topic does not exist (allow_auto_topic_creation).
    • The otelcol.receiver.kafka component has new configuration options max_partition_fetch_size, rack_id, and use_leader_epoch.
    • The otelcol.exporter.s3 component has new configuration options s3_base_prefix and s3_partition_timezone.
    • The otelcol.processor.servicegraph component now supports defining the maximum number of buckets for generated exponential histograms.
    • See the upstream [core][https://github.com/open-telemetry/opentelemetry-collector/blob/v0.139.0/CHANGELOG.md] and [contrib][https://github.com/open-telemetry/opentelemetry-collector-contrib/blob/v0.139.0/CHANGELOG.md] changelogs for more details.

Enhancements

  • Add per-application rate limiting with the strategy attribute in the faro.receiver component, to prevent one application from consuming the rate limit quota of others. (@hhertout)

  • Add support of tls in components loki.source.(awsfirehose|gcplog|heroku|api) and prometheus.receive_http and pyroscope.receive_http. (@fgouteroux)

  • Remove SendSIGKILL=no from unit files and recommendations (@oleg-kozlyuk-grafana)

  • Reduce memory overhead of prometheus.remote_write's WAL by lowering the size of the allocated series storage. (@kgeckhart)

  • Reduce lock wait/contention on the labelstore.LabelStore by removing unecessary usage from prometheus.relabel. (@kgeckhart)

  • prometheus.exporter.postgres dependency has been updated to v0.18.1. This includes new stat_progress_vacuum and buffercache_summary collectors, as well as other bugfixes and enhancements. (@cristiangreco)

  • Update Beyla component to 2.7.4. (@grcevski)

  • Support delimiters in stage.luhn. (@dehaansa)

  • pyroscope.java: update async-profiler to 4.2 (@korniltsev-grafanista)

  • Improve debug info output from exported receivers (loki, prometheus and pyroscope). (@kalleep)

  • prometheus.exporter.unix: Add an arp config block to configure the ARP collector. (@ptodev)

  • prometheus.exporter.snowflake dependency has been updated to 20251016132346-6d442402afb2, which updates data ownership queries to use last_over_time for a 24 hour period. (@dasomeone)

  • loki.source.podlogs now supports preserve_discovered_labels parameter to preserve discovered pod metadata labels for use by downstream components. (@QuentinBisson)

  • Rework underlying framework of Alloy UI to use Vite instead of Create React App. (@jharvey10)

  • Use POST requests for remote config requests to avoid hitting http2 header limits. (@tpaschalis)

  • loki.source.api during component shutdown will now reject all the inflight requests with status code 503 after graceful_shutdown_timeout has expired. (@kalleep)

  • kubernetes.discovery Add support for attaching namespace metadata. (@kgeckhart)

Bugfixes

  • Stop loki.source.kubernetes discarding log lines with duplicate timestamps. (@ciaranj)

  • Fix direction of arrows for pyroscope components in UI graph. (@dehaansa)

  • Only log EOF errors for syslog port investigations in loki.source.syslog as Debug, not Warn. (@dehaansa)

  • Fix prometheus.exporter.process ignoring the remove_empty_groups argument. (@mhamzahkhan)

  • Fix issues with "unknown series ref when trying to add exemplar" from prometheus.remote_write by allowing series ref links to be updated if they change. (@kgeckhart)

  • Fix loki.source.podlogs component to register the Kubernetes field index for spec.nodeName when node filtering is enabled, preventing "Index with name field:spec.nodeName does not exist" errors. (@QuentinBisson)

  • Fix issue in loki.source.file where scheduling files could take too long. (@kalleep)

  • Fix loki.write no longer includes internal labels __. (@matt-gp)

  • Fix missing native histograms custom buckets (NHCB) samples from prometheus.remote_write. (@krajorama)

  • otelcol.receiver.prometheus now supports mixed histograms if prometheus.scrape has honor_metadata set to true. (@ptodev) A mixed histogram is one which has both classic and exponential buckets.

  • loki.source.file has better support for non-UTF-8 encoded files. (@ptodev)

    • A BOM will be taken into account if the file is UTF-16 encoded and encoding is set to UTF-16. (Not UTF-16BE or UTF-16LE)
    • The carriage return symbol in Windows log files with CLRF endings will no longer be part of the log line.
    • These bugs used to cause some logs to show up with Chinese characters. Notably, this would happen on MSSQL UTF-16 LE logs.

  • Fix the loki.write endpoint block's enable_http2 attribute to actually affect the client. HTTP2 was previously disabled regardless of configuration. (@dehaansa)

  • Optionally remove trailing newlines before appending entries in stage.multiline. (@dehaansa)

v1.11.3

Enhancements

  • Schedule new path targets faster in loki.source.file. (@kalleep)

  • Add prometheus.static.exporter that exposes metrics specified in a text file in Prometheus exposition format. (@kalleep)

Bugfixes

  • local.file_match now publish targets faster whenever targets in arguments changes. (@kalleep)

  • Fix otelcol.exporter.splunkhec arguments missing documented otel_attrs_to_hec_metadata block. (@dehaansa)

  • Support Scrape Protocol specification in CRDS for prometheus.operator.* components. (@dehaansa)

  • Fix panic in otelcol.receiver.syslog when no tcp block was configured. (@kalleep)

  • Fix breaking changes in the texfile collector for prometheus.exporter.windows, and prometheus.exporter.unix, when prometheus/common was upgraded. (@kgeckhart)

  • Support recovering from corrupted positions file entries in loki.source.file. (@dehaansa)

Other changes

  • Augment prometheus.scrape 'scheme' argument strengthening link to protocol. (@lewismc)

  • Stop faro.receiver losing trace context when exception has stack trace. (@duartesaraiva98)

v1.11.2

Bugfixes

  • Fix potential deadlock in loki.source.journal when stopping or reloading the component. (@thampiotr)

  • Honor sync timeout when waiting for network availability for prometheus.operator.* components. (@dehaansa)

  • Fix prometheus.exporter.cloudwatch to not always emit debug logs but respect debug property. (@kalleep)

  • Fix an issue where component shutdown could block indefinitely by adding a warning log message and a deadline of 10 minutes. The deadline can be configured with the --feature.component-shutdown-deadline flag if the default is not suitable. (@thampiotr)

  • Fix potential deadlocks in loki.source.file and loki.source.journal when component is shutting down. (@kalleep, @thampiotr)

v1.11.0

Breaking changes

  • Prometheus dependency had a major version upgrade from v2.55.1 to v3.4.2. (@thampiotr)

    • The . pattern in regular expressions in PromQL matches newline characters now. With this change a regular expressions like .* matches strings that include \n. This applies to matchers in queries and relabel configs in Prometheus and Loki components.

    • The enable_http2 in prometheus.remote_write component's endpoints has been changed to false by default. Previously, in Prometheus v2 the remote write http client would default to use http2. In order to parallelize multiple remote write queues across multiple sockets its preferable to not default to http2. If you prefer to use http2 for remote write you must now set enable_http2 to true in your prometheus.remote_write endpoints configuration section.

    • The experimental CLI flag --feature.prometheus.metric-validation-scheme has been deprecated and has no effect. You can configure the metric validation scheme individually for each prometheus.scrape component.

    • Log message format has changed for some of the prometheus.* components as part of the upgrade to Prometheus v3.

    • The values of the le label of classic histograms and the quantile label of summaries are now normalized upon ingestion. In previous Alloy versions, that used Prometheus v2, the value of these labels depended on the scrape protocol (protobuf vs text format) in some situations. This led to label values changing based on the scrape protocol. E.g. a metric exposed as my_classic_hist{le="1"} would be ingested as my_classic_hist{le="1"} via the text format, but as my_classic_hist{le="1.0"} via protobuf. This changed the identity of the metric and caused problems when querying the metric. In current Alloy release, which uses Prometheus v3, these label values will always be normalized to a float like representation. I.e. the above example will always result in my_classic_hist{le="1.0"} being ingested into Prometheus, no matter via which protocol. The effect of this change is that alerts, recording rules and dashboards that directly reference label values as whole numbers such as le="1" will stop working.

      The recommended way to deal with this change is to fix references to integer le and quantile label values, but otherwise do nothing and accept that some queries that span the transition time will produce inaccurate or unexpected results.

    See the upstream Prometheus v3 migration guide for more details.

  • prometheus.exporter.windows dependency has been updated to v0.31.1. (@dehaansa)

    • There are various renamed metrics and two removed collectors (cs, logon), see the v1.11 release notes for more information.

  • scrape_native_histograms attribute for prometheus.scrape is now set to false, whereas in previous versions of Alloy it would default to true. This means that it is no longer enough to just configure scrape_protocols to start with PrometheusProto to scrape native histograms - scrape_native_histograms has to be enabled. If scrape_native_histograms is enabled, scrape_protocols will automatically be configured correctly for you to include PrometheusProto. If you configure it explicitly, Alloy will validate that PrometheusProto is in the scrape_protocols list.

  • Add otel_attrs_to_hec_metadata configuration block to otelcol.exporter.splunkhec to match otelcol.receiver.splunkhec. (@cgetzen)

  • [otelcol.processor.batch] Two arguments have different default values. (@ptodev)

    • send_batch_size is now set to 2000 by default. It used to be 8192.
    • send_batch_max_size is now set to 3000 by default. It used to be 0.
    • This helps prevent issues with ingestion of batches that are too large.

  • OpenTelemetry Collector dependencies upgraded from v0.128.0 to v0.134.0. (@ptodev)

    • The otelcol.receiver.opencensus component has been deprecated and will be removed in a future release, use otelcol.receiver.otelp instead.
    • [otelcol.exporter.*] The deprecated blocking argument in the sending_queue block has been removed. Use block_on_overflow instead.
    • [otelcol.receiver.kafka, otelcol.exporter.kafka]: Removed the broker_addr argument from the aws_msk block. Also removed the SASL/AWS_MSK_IAM authentication mechanism.
    • [otelcol.exporter.splunkhec] The batcher block is deprecated and will be removed in a future release. Use the queue block instead.
    • [otelcol.exporter.loadbalancing] Use a linear probe to decrease variance caused by hash collisions, which was causing a non-uniform distribution of loadbalancing.
    • [otelcol.connector.servicegraph] The database_name_attribute argument has been removed.
    • [otelcol.connector.spanmetrics] Adds a default maximum number of exemplars within the metric export interval.
    • [otelcol.processor.tail_sampling] Add a new block_on_overflow config attribute.

Features

  • Add the otelcol.receiver.fluentforward receiver to receive logs via Fluent Forward Protocol. (@rucciva)

  • Add the prometheus.enrich component to enrich metrics using labels from discovery.* components. (@ArkovKonstantin)

  • Add the otelcol.receiver.awsecscontainermetrics receiver (from upstream OTEL contrib) to read AWS ECS task- and container-level resource usage metrics. (@gregbrowndev)

  • Add node_filter configuration block to loki.source.podlogs component to enable node-based filtering for pod discovery. When enabled, only pods running on the specified node will be discovered and monitored, significantly reducing API server load and network traffic in DaemonSet deployments. (@QuentinBisson)

  • (Experimental) Additions to experimental database_observability.mysql component:

    • query_sample collector now supports auto-enabling the necessary setup_consumers settings (@cristiangreco)
    • query_sample collector is now compatible with mysql less than 8.0.28 (@cristiangreco)
    • include server_id label on log entries (@matthewnolf)
    • support receiving targets argument and relabel those to include server_id (@matthewnolf)
    • updated the config blocks and documentation (@cristiangreco)

  • (Experimental) Additions to experimental database_observability.postgres component:

    • add query_tables collector for postgres (@matthewnolf)
    • add cloud_provider.aws configuration that enables optionally supplying the ARN of the database under observation. The ARN is appended to metric samples as labels for easier filtering and grouping of resources.
    • add query_sample collector for postgres (@gaantunes)
    • add schema_details collector for postgres (@fridgepoet)
    • include server_id label on logs and metrics (@matthewnolf)

  • Add otelcol.receiver.googlecloudpubsub community component to receive metrics, traces, and logs from Google Cloud Pub/Sub subscription. (@eraac)

  • Add otel collector converter for otelcol.receiver.googlecloudpubsub. (@kalleep)

  • (Experimental) Add a honor_metadata configuration argument to the prometheus.scrape component. When set to true, it will propagate metric metadata to downstream components.

  • Add a flag to pyroscope.ebpf alloy configuration to set the off-cpu profiling threshold. (@luweglarz)

  • Add encoding.url_encode and encoding.url_decode std lib functions. (@kalleep)

Enhancements

  • Ensure text in the UI does not overflow node boundaries in the graph. (@blewis12)

  • Fix pyroscope.write component's AppendIngest method to respect configured timeout and implement retry logic. The method now properly uses the configured remote_timeout, includes retry logic with exponential backoff, and tracks metrics for sent/dropped bytes and profiles consistently with the Append method. (@korniltsev)

  • pyroscope.write, pyroscope.receive_http components include trace_id in logs and propagate it downstream. (@korniltsev)

  • Improve logging in pyroscope.write component. (@korniltsev)

  • Add comprehensive latency metrics to pyroscope.write component with endpoint-specific tracking for both push and ingest operations. (@korniltsev, @claude)

  • prometheus.scrape now supports convert_classic_histograms_to_nhcb, enable_compression, metric_name_validation_scheme, metric_name_escaping_scheme, native_histogram_bucket_limit, and native_histogram_min_bucket_factor arguments. See reference documentation for more details. (@thampiotr)

  • Add max_send_message_size configuration option to loki.source.api component to control the maximum size of requests to the push API. (@thampiotr)

  • Add protobuf_message argument to prometheus.remote_write endpoint configuration to support both Prometheus Remote Write v1 and v2 protocols. The default remains "prometheus.WriteRequest" (v1) for backward compatibility. (@thampiotr)

  • Update the yet-another-cloudwatch-exporter dependency to point to the prometheus-community repo as it has been donated. Adds a few new services to prometheus.exporter.cloudwatch. (@dehaansa, @BoweFlex, @andriikushch)

  • pyroscope.java now supports configuring the log_level and quiet flags on async-profiler. (@deltamualpha)

  • Add application_host and network_inter_zone features to beyla.ebpf component. (@marctc)

  • Set the publisher name in the Windows installer to "Grafana Labs". (@martincostello)

  • Switch to the community maintained fork of go-jmespath that has more features. (@dehaansa)

  • Add a stage.pattern stage to loki.process that uses LogQL patterns to parse logs. (@dehaansa)

  • Add support to validate references, stdlib functions and arguments when using validate command. (@kalleep)

  • Update the prometheus.exporter.process component to get the remove_empty_groups option. (@dehaansa)

  • Remove unnecessary allocations in stage.static_labels. (@kalleep)

  • Upgrade beyla.ebpf from Beyla version v2.2.5 to v2.5.8 The full list of changes can be found in the Beyla release notes (@marctc)

  • prometheus.exporter.azure supports setting interval and timespan independently allowing for further look back when querying metrics. (@kgeckhart)

  • loki.source.journal now supports legacy_positon block that can be used to translate Static Agent or Promtail position files. (@kalleep)

  • Normalize attr key name in logfmt logger. (@zry98)

  • (Experimental) Add an extra parameter to the array.combine_maps standard library function to enable preserving the first input list even if there is no match. (@ptodev)

  • Reduce memory overhead of prometheus.remote_write's WAL by bringing in an upstream change to only track series in a slice if there's a hash conflict. (@kgeckhart)

  • Reduce log level from warning for loki.write when request fails and will be retried. (@kalleep)

  • Fix slow updates to loki.source.file when only targets have changed and pipeline is blocked on writes. (@kalleep)

  • Reduced allocation in loki.write when using external labels with mutliple endpoints. (@kalleep)

  • The Windows installer and executables are now code signed. (@martincostello)

  • Reduce compressed request size in prometheus.write.queue by ensuring append order is maintained when sending metrics to the WAL. (@kgeckhart)

  • Add protobuf_message and metadata_cache_size arguments to prometheus.write.queue endpoint configuration to support both Prometheus Remote Write v1 and v2 protocols. The default remains "prometheus.WriteRequest" (v1) for backward compatibility. (@dehaansa)

  • Reduce allocations for loki.process when stage.template is used. (@kalleep)

  • Reduce CPU of prometheus.write.queue by eliminating duplicate calls to calculate the protobuf Size. (@kgeckhart)

  • Use new cache for metadata cache in prometheus.write.queue and support disabling the metadata cache with it disable by default. (@kgeckhart, @dehaansa)

Bugfixes

  • Update webdevops/go-common dependency to resolve concurrent map write panic. (@dehaansa)

  • Fix ebpf profiler metrics pyroscope_ebpf_active_targets, pyroscope_ebpf_profiling_sessions_total, pyroscope_ebpf_profiling_sessions_failing_total not being updated. (luweglarz)

  • Fix prometheus.operator.podmonitors so it now handle portNumber from PodMonitor CRD. (@kalleep)

  • Fix pyroscope.receive_http so it does not restart server if the server configuration has not changed. (@korniltsev)

  • Increase default connection limit in pyroscope.receive_http from 100 to 16k. (@korniltsev)

  • Fix issue in prometheus.remote_write's WAL which could allow it to hold an active series forever. (@kgeckhart)

  • Fix issue in static and promtail converter where metrics type was not properly handled. (@kalleep)

  • Fix prometheus.operator.* components to allow them to scrape correctly Prometheus Operator CRDs. (@thomas-gouveia)

  • Fix database_observability.mysql and database_observability.postgres crashing alloy process due to uncaught errors.

  • Fix data race inloki.source.docker that could cause Alloy to panic. (@kalleep)

  • Fix race conditions in loki.source.syslog where it could deadlock or cause port bind errors during config reload or shutdown. (@thampiotr)

  • Fix prometheus.exporter.redis component so that it no longer ignores the MaxDistinctKeyGroups configuration option. If key group metrics are enabled, this will increase the cardinality of the generated metrics. (@stegosaurus21)

  • Fix loki.source.podlogs component to properly collect logs from Kubernetes Jobs and CronJobs. Previously, the component would fail to scrape logs from short-lived or terminated jobs due to race conditions between job completion and pod discovery. The fix includes:

    • Job-aware termination logic with extended grace periods (10-60 seconds) to ensure all logs are captured
    • Proper handling of pod deletion and race conditions between job completion and controller cleanup
    • Separation of concerns: shouldStopTailingContainer() handles standard Kubernetes restart policies for regular pods, while shouldStopTailingJobContainer() handles job-specific lifecycle with grace periods
    • Enhanced deduplication mechanisms to prevent duplicate log collection while ensuring comprehensive coverage
    • Comprehensive test coverage including unit tests and deduplication validation This resolves the issue where job logs were being missed, particularly for fast-completing jobs or jobs that terminated before discovery. (@QuentinBisson)

  • Fix loki.source.journal creation failing with an error when the journal file is not found. (@thampiotr)

  • Fix graph UI so it generates correct URLs for components in remotecfg modules. (@patrickeasters)

  • Fix panic in loki.write when component is shutting down and external_labels are configured. (@kalleep)

  • Fix excessive debug logs always being emitted by prometheus.exporter.mongodb. (@kalleep)

v1.10.2

Bugfixes

  • Fix issue in prometheus.write.queue causing inability to increase shard count if existing WAL data was present on start. (@kgeckhart)

  • Fix issue with loki.source.gcplog when push messages sent by gcp pub/sub only includes messageId. (@kalleep)

v1.10.1

Bugfixes

  • Fix issue with faro.receiver cors not allowing X-Scope-OrgID and traceparent headers. (@mar4uk)

  • Fix issues with propagating cluster peers change notifications to components configured with remotecfg. (@dehaansa)

  • Fix issues with statistics reporter not including components only configured with remotecfg. (@dehaansa)

  • Fix issues with prometheus.exporter.windows not propagating dns collector config. (@dehaansa)

  • Fixed a bug in prometheus.write.queue which caused retries even when max_retry_attempts was set to 0. (@ptodev)

  • Fixed a bug in prometheus.write.queue which caused labelling issues when providing more than one label in external_labels. (@dehaansa)

  • Add application_host and network_inter_zone features to beyla.ebpf component. (@marctc)

  • Fix issues in loki.process where stage.multiline did not pass through structured metadata. (@jan-mrm)

  • Fix URLs in the Windows installer being wrapped in quotes. (@martincostello)

  • Fixed an issue where certain otelcol.* components could prevent Alloy from shutting down when provided invalid configuration. (@thampiotr)

v1.10.0

Breaking changes

  • Removing the nanoserver-1809 container image for Windows 2019. (@ptodev) This is due to the deprecation of windows-2019 GitHub Actions runners. The windowsservercore-ltsc2022 Alloy image is still being published to DockerHub.

Bugfixes

  • Upgrade otelcol components from OpenTelemetry v0.126.0 to v0.128.0 (@korniltsev, @dehaansa)
    • [otelcol.exporter.kafka]: Allow kafka exporter to produce to topics based on metadata key values.
    • [otelcol.receiver.kafka]: Enforce a backoff mechanism on non-permanent errors, such as when the queue is full.
    • [otelcol.receiver.kafka]: Don't restart the Kafka consumer on failed errors when message marking is enabled for them.
    • [otelcol.exporter.datadog]: Fix automatic intial point dropping when converting cumulative monotonic sum metrics.
    • [otelcol.exporter.datadog]: config tls::insecure_skip_verify is now taken into account in metrics path.
    • [otelcol.exporter.datadog]: Correctly treat summary counts as cumulative monotonic sums instead of cumulative non-monotonic sums.
    • [otelcol.connector.spanmetrics]: Fix bug causing span metrics calls count to be always 0 when using delta temporality.
    • [otelcol.exporter.splunkhec]: Treat HTTP 403 Forbidden as a permanent error.

Features

  • (Experimental) Add an array.group_by stdlib function to group items in an array by a key. (@wildum)

  • Add the otelcol.exporter.faro exporter to export traces and logs to Faro endpoint. (@mar4uk)

  • Add the otelcol.receiver.faro receiver to receive traces and logs from the Grafana Faro Web SDK. (@mar4uk)

  • Add entropy support for loki.secretfilter (@romain-gaillard)

Enhancements

  • Add hash_string_id argument to foreach block to hash the string representation of the pipeline id instead of using the string itself. (@wildum)

  • Update async-profiler binaries for pyroscope.java to 4.0-87b7b42 (@github-hamza-bouqal)

  • (Experimental) Additions to experimental database_observability.mysql component:

    • Add explain_plan collector to database_observability.mysql component. (@rgeyer)
    • locks: addition of data locks collector (@gaantunes @fridgepoet)
    • query_sample collector is now enabled by default (@matthewnolf)
    • query_tables collector now deals better with truncated statements (@cristiangreco)

  • (Experimental) prometheus.write.queue add support for exemplars. (@dehaansa)

  • (Experimental) prometheus.write.queue initialize queue metrics that are seconds values as time.Now, not 0. (@dehaansa)

  • Update secret-filter gitleaks.toml from v8.19.0 to v8.26.0 (@andrejshapal)

  • Wire in survey block for beyla.ebpf component. (@grcevski, @tpaschalis)

  • Upgrade otelcol components from OpenTelemetry v0.126.0 to v0.128.0 (@korniltsev, @dehaansa)

    • [otelcol.processor.resourcedetection]: Add additional OS properties to resource detection: os.build.id and os.name.
    • [otelcol.processor.resourcedetection]: Add host.interface resource attribute to system detector.
    • [otelcol.exporter.kafka]: Fix Snappy compression codec support for the Kafka exporter.
    • [otelcol.receiver.filelog]: Introduce utf8-raw encoding to avoid replacing invalid bytes with \uFFFD when reading UTF-8 input.
    • [otelcol.processor.k8sattributes]: Support extracting labels and annotations from k8s Deployments.
    • [otelcol.processor.k8sattributes]: Add option to configure automatic service resource attributes.
    • [otelcol.exporter.datadog]: Adds hostname_detection_timeout configuration option for Datadog Exporter and sets default to 25 seconds.
    • [otelcol.receiver.datadog]: Address semantic conventions noncompliance and add support for http/db.
    • [otelcol.exporter.awss3]: Add the retry mode, max attempts and max backoff to the settings.

  • Add enable_tracing attribute to prometheus.exporter.snowflake component to support debugging issues. (@dehaansa)

  • Add support for conditions and statement-specific error_mode in otelcol.processor.transform. (@ptodev)

  • Add storage and start_from args to cloudwatch logs receiver. (@boernd)

  • Reduced allocation in Loki processing pipelines. (@thampiotr)

  • Update the prometheus.exporter.postgres component with latest changes and bugfixes for Postgres17 (@cristiangreco)

  • Add tail_from_end argument to loki.source.podlogs to optionally start reading from the end of a log stream for newly discovered pods. (@harshrai654)

  • Remove limitation in loki.source.file when legacy_position_file is unset. Alloy can now recover legacy positions even if labels are added. (@kalleep)

Bugfixes

  • Fix path for correct injection of version into constants at build time. (@adlotsof)

  • Propagate the -feature.community-components.enabled flag for remote configuration components. (@tpaschalis)

  • Fix extension registration for otelcol.receiver.splunkhec auth extensions. (@dehaansa)

Other changes

  • Mark pyroscope.receive_http and pyroscope.relabel components as GA. (@marcsanmi)

  • Upgrade otelcol.exporter.windows to v0.30.8 to get bugfixes and fix update collector support. (@dehaansa)

  • Add User-Agent header to remotecfg requests. (@tpaschalis)

v1.9.2

Bugfixes

  • Send profiles concurrently from pyroscope.ebpf. (@korniltsev)

  • Fix the validate command not understanding the livedebugging block. (@dehaansa)

  • Fix invalid class names in python profiles obtained with pyroscope.ebpf. (@korniltsev)

  • Fixed a bug which prevented non-secret optional secrets to be passed in as number arguments. (@ptodev)

  • For CRD-based components (prometheus.operator.*), retry initializing informers if the apiserver request fails. This rectifies issues where the apiserver is not reachable immediately after node restart. (@dehaansa)

Other changes

  • Add no-op blocks and attributes to the prometheus.exporter.windows component (@ptodev). Version 1.9.0 of Alloy removed the msmq block, as well as the enable_v2_collector, where_clause, and use_api attributes in the service block. This made it difficult for users to upgrade, so those attributes have now been made a no-op instead of being removed.

v1.9.1

Features

  • Update the prometheus.exporter.windows component to version v0.30.7. This adds new metrics to the dns collector. (@dehaansa)

Bugfixes

  • Update the prometheus.exporter.windows component to version v0.30.7. This fixes an error with the exchange collector and terminal_services collector (@dehaansa)

  • Fix loki.source.firehose to propagate specific cloudwatch event timestamps when useIncomingTs is set to true. (@michaelPotter)

  • Fix elevated CPU usage when using some otelcol components due to debug logging. (@thampiotr)

Other changes

  • Upgrade otelcol components from OpenTelemetry v0.125.0 to v0.126.0 (@dehaansa):
    • [pkg/ottl] Add support for HasPrefix and HasSuffix functions.
    • [pkg/configtls] Add trusted platform module (TPM) support to TLS authentication for all otelcol components supporting TLS.
    • [otelcol.connector.spanmetrics] Add calls_dimension and histogram:dimension blocks for configuring additional dimensions for traces.span.metrics.calls and traces.span.metrics.duration metrics.
    • [otelcol.exporter.datadog] Enable instrumentation_scope_metadata_as_tags by default.
    • [otelcol.exporter.kafka] support configuration of compression level in producer configuration.
    • [otelcol.processor.tailsampling] invert sample and inverted not sample decisions deprecated, use the drop policy instead to explicitly not sample traces.
    • [otelcol.receiver.filelog] support compression value of auto to automatically detect file compression type.

v1.9.0

Breaking changes

  • The prometheus.exporter.windows component has been update to version v0.30.6. This update includes a significant rework of the exporter and includes some breaking changes. (@dehaansa)

    • The msmq and service collectors can no longer be configured with a WMI where clause. Any filtering previously done in a where clause will need to be done in a prometheus.relabel component.
    • The service collector no longer provides enable_v2_collector and use_api configuration options.
    • The mscluster_* and netframework_* collectors are now replaced with one mscluster and netframework collector that allows you to enable the separate metric groupings individually.
    • The teradici_pcoip and vmware_blast collectors have been removed from the exporter.

  • The prometheus.exporter.oracledb component now embeds the oracledb_exporter from oracle instead of the deprecated oracledb_exporter from iamseth for collecting metrics from an OracleDB server: (@wildum)

    • The arguments username, password, default_metrics, and custom_metrics are now supported.
    • The previously undocumented argument custom_metrics is now expecting a list of paths to custom metrics files.
    • The following metrics are no longer available by default: oracledb_sessions_activity, oracledb_tablespace_free_bytes

  • (Experimental) The enable_context_propagation argument in beyla.ebpf has been replaced with the context_propagation argument. Set enable_context_propagation to all to get the same behaviour as enable_context_propagation being set to true.

Features

  • Bump snmp_exporter and embedded modules in prometheus.exporter.snmp to v0.29.0, add cisco_device module support (@v-zhuravlev)

  • Add the otelcol.storage.file extension to support persistent sending queues and otelcol.receiver.filelog file state tracking between restarts. (@dehaansa)

  • Add otelcol.exporter.googlecloud community component to export metrics, traces, and logs to Google Cloud. (@motoki317)

  • Add support to configure basic authentication for alloy http server. (@kalleep)

  • Add validate command to alloy that will perform limited validation of alloy configuration files. (@kalleep)

  • Add support to validate foreach block when using validate command. (@kalleep)

  • Add otelcol.receiver.splunkhec component to receive events in splunk hec format and forward them to other otelcol.* components. (@kalleep)

  • Add support for Mimir federated rule groups in mimir.rules.kubernetes (@QuentinBisson)

Enhancements

  • prometheus.exporter.windows has been significantly refactored upstream and includes new collectors like filetime, pagefile, performancecounter, udp, and update as well as new configuration options for existing collectors. (@dehaansa)

  • prometheus.exporter.mongodb now offers fine-grained control over collected metrics with new configuration options. (@TeTeHacko)

  • Add binary version to constants exposed in configuration file syntatx. (@adlots)

  • Update loki.secretfilter to include metrics about redactions (@kelnage)

  • (Experimental) Various changes to the experimental component database_observability.mysql:

    • schema_table: add support for index expressions (@cristiangreco)
    • query_sample: enable opt-in support to extract unredacted sql query (sql_text) (@matthewnolf)
    • query_tables: improve queries parsing (@cristiangreco)
    • make tidbparser the default choice (@cristiangreco)
    • query_sample: better handling of timer overflows (@fridgepoet)
    • collect metrics on enabled performance_schema.setup_consumers (@fridgepoet)
    • query_sample: base log entries on calculated timestamp from rows, not now() (@fridgepoet)
    • query_sample: check digest is not null (@cristiangreco)
    • query_sample: add additional logs for wait events (@fridgepoet)
    • make tidb the default and only sql parser

  • Mixin dashboards improvements: added minimum cluster size to Cluster Overview dashboard, fixed units in OpenTelemetry dashboard, fixed slow components evaluation time units in Controller dashboard and updated Prometheus dashboard to correctly aggregate across instances. (@thampiotr)

  • Reduced the lag time during targets handover in a cluster in prometheus.scrape components by reducing thread contention. (@thampiotr)

  • Pretty print diagnostic errors when using alloy run (@kalleep)

  • Add labels_from_groups attribute to stage.regex in loki.process to automatically add named capture groups as labels. (@harshrai654)

  • The loki.rules.kubernetes component now supports adding extra label matchers to all queries discovered via PrometheusRule CRDs. (@QuentinBisson)

  • Add optional id field to foreach block to generate more meaningful component paths in metrics by using a specific field from collection items. (@harshrai654)

  • The mimir.rules.kubernetes component now supports adding extra label matchers to all queries discovered via PrometheusRule CRDs by extracting label values defined on the PrometheusRule. (@QuentinBisson)

  • Fix validation logic in beyla.ebpf component to ensure that either metrics or traces are enabled. (@marctc)

  • Improve foreach UI and add graph support for it. (@wildum)

  • Update statsd_exporter to v0.28.0, most notable changes: (@kalleep)

    • [0.23.0] Support experimental native histograms.
    • [0.24.1] Support scaling parameter in mapping.
    • [0.26.0] Add option to honor original labels from event tags over labels specified in mapping configuration.
    • [0.27.1] Support dogstatsd extended aggregation
    • [0.27.2] Fix panic on certain invalid lines

  • Upgrade beyla.ebpf to v2.2.4-alloy. The full list of changes can be found in the Beyla release notes. (@grcevski)

Bugfixes

  • Fix otelcol.receiver.filelog documentation's default value for start_at. (@petewall)

  • Fix pyroscope.scrape scraping godeltaprof profiles. (@korniltsev)

  • Fix #3386 lower casing scheme in prometheus.operator.scrapeconfigs. (@alex-berger)

  • Fix #3437 Component Graph links now follow --server.http.ui-path-prefix. (@solidcellaMoon)

  • Fix a bug in the foreach preventing the UI from showing the components in the template when the block was re-evaluated. (@wildum)

  • Fix alloy health handler so header is written before response body. (@kalleep)

  • Fix prometheus.exporter.unix to pass hwmon config correctly. (@kalleep)

  • Fix #3408 loki.source.docker can now collect logs from containers not in the running state. (@adamamsmith)

Other changes

  • Update the zap logging adapter used by otelcol components to log arrays and objects. (@dehaansa)

  • Updated Windows install script to add DisplayVersion into registry on install (@enessene)

  • Update Docker builds to install latest Linux security fixes on top of base image (@jharvey10)

  • Reduce Docker image size slightly by consolidating some RUN layers (@AchimGrolimund)

  • RPM artifacts in Alloy GitHub releases are no longer signed. The artifacts on the https://rpm.grafana.com repository used by the yum package manager will continue to be signed. (@ptodev)

  • Upgrade otelcol components from OpenTelemetry v0.122.0 to v0.125.0 (@ptodev):

    • [pkg/ottl] Enhance the Decode OTTL function to support all flavors of Base64.
    • [otelcol.processor.resourcedetection] Adding the os.version resource attribute to system processor.
    • [otelcol.auth.bearer] Allow the header name to be customized.
    • [otelcol.exporter.awss3] Add a new sending_queue feature.
    • [otelcol.exporter.awss3] Add a new timeout argument.
    • [otelcol.exporter.awss3] Add a new resource_attrs_to_s3 configuration block.
    • [otelcol.exporter.awss3] Fixes an issue where the AWS S3 Exporter was forcing an ACL to be set, leading to unexpected behavior in S3 bucket permissions.
    • [otelcol.connector.spanmetrics] A new include_instrumentation_scope configuration argument.
    • [otelcol.connector.spanmetrics] Initialise new calls_total metrics at 0.
    • [otelcol.connector.spanmetrics] A new aggregation_cardinality_limit configuration argument to limit the number of unique combinations of dimensions that will be tracked for metrics aggregation.
    • [otelcol.connector.spanmetrics] Deprecate the unused argument dimensions_cache_size.
    • [otelcol.connector.spanmetrics] Moving the start timestamp (and last seen timestamp) from the resourceMetrics level to the individual metrics level. This will ensure that each metric has its own accurate start and last seen timestamps, regardless of its relationship to other spans.
    • [otelcol.processor.k8sattributes] Add option to configure automatic resource attributes - with annotation prefix. Implements Specify resource attributes using Kubernetes annotations.
    • [otelcol.connector.servicegraph] Change database_name_attribute to accept a list of values.
    • [otelcol.exporter.kafka, otelcol.receiver.kafka] Deprecating the auth > plain_text block. Use auth > sasl with mechanism set to PLAIN instead.
    • [otelcol.exporter.kafka, otelcol.receiver.kafka] Deprecating the topic argument. Use logs > topic, metrics > topic, or traces > topic instead.
    • [otelcol.exporter.kafka, otelcol.receiver.kafka] Deprecate the auth > tls block. Use the top-level tls block instead.
    • [otelcol.receiver.kafka] Add max_fetch_wait config setting. This setting allows you to specify the maximum time that the broker will wait for min_fetch_size bytes of data to be available before sending a response to the client.
    • [ otelcol.receiver.kafka] Add support for configuring Kafka consumer rebalance strategy and group instance ID.

v1.8.3

Bugfixes

  • Fix mimir.rules.kubernetes panic on non-leader debug info retrieval (@TheoBrigitte)

  • Fix detection of the "streams limit exceeded" error in the Loki client so that metrics are correctly labeled as ReasonStreamLimited. (@maratkhv)

  • Fix loki.source.file race condition that often lead to panic when using decompression. (@kalleep)

  • Fix deadlock in loki.source.file that can happen when targets are removed. (@kalleep)

  • Fix loki.process to emit valid logfmt. (@kalleep)

v1.8.2

Bugfixes

  • Fix otelcol.exporter.prometheus dropping valid exemplars. (@github-vincent-miszczak)

  • Fix loki.source.podlogs not adding labels __meta_kubernetes_namespace and __meta_kubernetes_pod_label_*. (@kalleep)

v1.8.1

Bugfixes

  • rfc3164_default_to_current_year argument was not fully added to loki.source.syslog (@dehaansa)

  • Fix issue with remoteCfg service stopping immediately and logging noop error if not configured (@dehaansa)

  • Fix potential race condition in remoteCfg service metrics registration (@kalleep)

  • Fix panic in prometheus.exporter.postgres when using minimal url as data source name. (@kalleep)

v1.8.0

Breaking changes

  • Removed open_port and executable_name from top level configuration of Beyla component. Removed enabled argument from network block. (@marctc)

  • Breaking changes from the OpenTelemetry Collector v0.122 update: (@wildum)

    • otelcol.exporter.splunkhec: min_size_items and max_size_items were replaced by min_size, max_size and sizer in the batcher block to allow users to configure the size of the batch in a more flexible way.
    • The telemetry level of Otel components is no longer configurable. The level argument in the debug_metrics block is kept to avoid breaking changes but it is not used anymore.
    • otelcol.processor.tailsampling changed the unit of the decision timer metric from microseconds to milliseconds. (change unit of otelcol_processor_tail_sampling_sampling_decision_timer_latency)
    • otelcol.processor.deltatocumulative: rename otelcol_deltatocumulative_datapoints_processed to otelcol_deltatocumulative_datapoints and remove the metrics otelcol_deltatocumulative_streams_evicted, otelcol_deltatocumulative_datapoints_dropped and otelcol_deltatocumulative_gaps_length.
    • The regex attribute was removed from otelcol.processor.k8sattributes. The extract-patterns function from otelcol.processor.transform can be used instead.
    • The default value of metrics_flush_interval in otelcol.connector.servicegraph was changed from 0s to 60s.
    • s3_partition in otelcol.exporter.awss3 was replaced by s3_partition_format.

  • (Experimental) prometheus.write.queue metric names changed to align better with prometheus standards. (@mattdurham)

Features

  • Add otelcol.receiver.awscloudwatch component to receive logs from AWS CloudWatch and forward them to other otelcol.* components. (@wildum)
  • Add loki.enrich component to enrich logs using labels from discovery.* components. (@v-zhuravlev)
  • Add string concatenation for secrets type (@ravishankar15)
  • Add support for environment variables to OpenTelemetry Collector config. (@jharvey10)
  • Replace graph in Alloy UI with a new version that supports modules and data flow visualization. (@wildum)
  • Added --cluster.wait-for-size and --cluster.wait-timeout flags which allow to specify the minimum cluster size required before components that use clustering begin processing traffic to ensure adequate cluster capacity is available. (@thampiotr)
  • Add trace_printer to beyla.ebpf component to print trace information in a specific format. (@marctc)
  • Add support for live debugging and graph in the UI for components imported via remotecfg. (@wildum)

Enhancements

  • Add the ability to set user for Windows Service with silent install (@dehaansa)

  • Add livedebugging support for structured_metadata in loki.process (@dehaansa)

  • (Public Preview) Add a --windows.priority flag to the run command, allowing users to set windows process priority for Alloy. (@dehaansa)

  • (Experimental) Adding a new prometheus.operator.scrapeconfigs which discovers and scrapes ScrapeConfig Kubernetes resources. (@alex-berger)

  • Add rfc3164_default_to_current_year argument to loki.source.syslog (@dehaansa)

  • Add connection_name support for prometheus.exporter.mssql (@bck01215)

  • Add livedebugging support for prometheus.scrape (@ravishankar15, @wildum)

  • Have loki.echo log the entry_timestamp and structured_metadata for any loki entries received (@dehaansa)

  • Bump snmp_exporter and embedded modules in prometheus.exporter.snmp to v0.28.0 (@v-zhuravlev)

  • Update mysqld_exporter to v0.17.2, most notable changes: (@cristiangreco)

    • [0.17.1] Add perf_schema quantile columns to collector
    • [0.17.1] Fix database quoting problem in collector 'info_schema.tables'
    • [0.17.1] Use SUM_LOCK_TIME and SUM_CPU_TIME with mysql >= 8.0.28
    • [0.17.1] Fix query on perf_schema.events_statements_summary_by_digest
    • [0.17.2] Fix query on events_statements_summary_by_digest for mariadb

  • Added additional backwards compatibility metrics to prometheus.write.queue. (@mattdurham)

  • Add new stdlib functions encoding.to_json (@ravishankar15)

  • Added OpenTelemetry logs and metrics support to Alloy mixin's dashboards and alerts. (@thampiotr)

  • Add support for proxy and headers in prometheus.write.queue. (@mattdurham)

  • Added support for switching namespace between authentication and kv retrieval to support Vault Enterprise (@notedop)

  • (Experimental) Various changes to the experimental component database_observability.mysql:

    • query_sample: better handling of truncated queries (@cristiangreco)
    • query_sample: add option to use TiDB sql parser (@cristiangreco)
    • query_tables: rename collector from query_sample to better reflect responsibility (@matthewnolf)
    • query_sample: add new collector that replaces previous implementation to collect more detailed sample information (@matthewnolf)
    • query_sample: refactor parsing of truncated queries (@cristiangreco)

  • Add labels validation in pyroscope.write to prevent duplicate labels and invalid label names/values. (@marcsanmi)

  • Reduced lock contention in prometheus.scrape component (@thampiotr)

  • Support converting otel config which uses a common receiver across pipelines with different names. (@wildum)

  • Reduce CPU usage of the loki.source.podlogs component when pods logs target lots of pods (@QuentinBisson)

  • Add error body propagation in pyroscope.write, for /ingest calls. (@simonswine)

  • Add tenant label to remaining loki_write_.+ metrics (@towolf)

  • Removed syntax highlighting from the component details UI view to improve rendering performance. (@tpaschalis)

  • A new grafana/alloy:vX.Y.Z-windowsservercore-ltsc2022 Docker image is now published on DockerHub. (@ptodev)

Bugfixes

  • Fix deadlocks in loki.source.file when tailing fails (@mblaschke)

  • Add missing RBAC permission for ScrapeConfig (@alex-berger)

  • Fixed an issue in the mimir.rules.kubernetes component that would keep the component as unhealthy even when it managed to start after temporary errors (@nicolasvan)

  • Allow kafka exporter to attempt to connect even if TLS enabled but cert & key are not specified (@dehaansa)

  • Fixed bug where all resources were not being collected from prometheus.exporter.azure when using regions (@kgeckhart)

  • Fix panic in loki.source.file when the tailer had no time to run before the runner was stopped (@wildum)

Other changes

  • Upgrading to Prometheus v2.55.1. (@ptodev)

    • Added a new http_headers argument to many discovery and prometheus components.
    • Added a new scrape_failure_log_file argument to prometheus.scrape.

  • Non-breaking changes from the OpenTelemetry Collector v0.122 update: (@wildum)

    • otelcol.processor.transform has a new statements block for transformations which don't require a context to be specified explicitly.
    • otelcol.receiver.syslog has a new on_error argument to specify the action to take when an error occurs while receiving logs.
    • otelcol.processor.resourcedetection now supports dynatrace as a resource detector.
    • otelcol.receiver.kafka has a new error_backoff block to configure how failed requests are retried.
    • otelcol.receiver.vcenter has three new metrics vcenter.vm.cpu.time, vcenter.vm.network.broadcast.packet.rate and vcenter.vm.network.multicast.packet.rate.
    • otelcol.exporter.awss3 has two new arguments acl and storage_class.
    • otelcol.auth.headers headers can now be populated using Authentication metadata using from_attribute

  • Change the stability of the beyla.ebpf component from "public preview" to "generally available". (@marctc)

  • The ingest API of pyroscope.receive_http no longer forwards all received headers, instead only passes through the Content-Type header. (@simonswine)

v1.7.5

Enhancements

  • Set zstd as default compression for prometheus.write.queue. (@mattdurham)

v1.7.4

Bugfixes

  • Revert the changes to loki.source.file from release v1.7.0. These changes introduced a potential deadlock. (@dehaansa)

v1.7.3

Breaking changes

  • Fixed the parsing of selections, application and network filter blocks for Beyla. (@raffaelroquetto)

Enhancements

  • Add the stat_checkpointer collector in prometheus.exporter.postgres (@dehaansa)

Bugfixes

  • Update the prometheus.exporter.postgres component to correctly support Postgres17 when stat_bgwriter collector is enabled (@dehaansa)

  • Fix remoteCfg logging and metrics reporting of errNotModified as a failure (@zackman0010)

v1.7.2

Bugfixes

  • Fixed an issue where the otelcol.exporter.awss3 could not be started with the sumo_ic marshaler. (@wildum)

  • Update jfr-parser dependency to v0.9.3 to fix jfr parsing issues in pyroscope.java. (@korniltsev)

  • Fixed an issue where passing targets from some standard library functions was failing with target::ConvertFrom error. (@thampiotr)

  • Fixed an issue where indexing targets as maps (e.g. target["foo"]) or objects (e.g. target.foo) or using them with certain standard library functions was resulting in expected object or array, got capsule error under some circumstances. This could also lead to foreach evaluation failed errors when using the foreach configuration block. (@thampiotr)

  • Update prometheus.write.queue to reduce memory fragmentation and increase sent throughput. (@mattdurham)

  • Fixed an issue where the otelcol.exporter.kafka component would not start if the encoding was specific to a signal type. (@wildum)

v1.7.1

Bugfixes

  • Fixed an issue where some exporters such as prometheus.exporter.snmp couldn't accept targets from other components with an error conversion to '*map[string]string' is not supported". (@thampiotr)

  • Enable batching of calls to the appender in prometheus.write.queue to reduce lock contention when scraping, which will lead to reduced scrape duration. (@mattdurham)

v1.7.0

Breaking changes

  • (Experimental) In prometheus.write.queue changed parallelism from attribute to a block to allow for dynamic scaling. (@mattdurham)

  • Remove tls_basic_auth_config_path attribute from prometheus.exporter.mongodb configuration as it does not configure TLS client behavior as previously documented.

  • Remove encoding and encoding_file_ext from otelcol.exporter.awss3 component as it was not wired in to the otel component and Alloy does not currently integrate the upstream encoding extensions that this would utilize.

Features

  • Add a otelcol.receiver.tcplog component to receive OpenTelemetry logs over a TCP connection. (@nosammai)

  • (Public preview) Add otelcol.receiver.filelog component to read otel log entries from files (@dehaansa)

  • (Public preview) Add a otelcol.processor.cumulativetodelta component to convert metrics from cumulative temporality to delta. (@madaraszg-tulip)

  • (Experimental) Add a stage.windowsevent block in the loki.process component. This aims to replace the existing stage.eventlogmessage. (@wildum)

  • Add pyroscope.relabel component to modify or filter profiles using Prometheus relabeling rules. (@marcsanmi)

  • (Experimental) A new foreach block which starts an Alloy pipeline for each item inside a list. (@wildum, @thampiotr, @ptodev)

Enhancements

  • Upgrade to OpenTelemetry Collector v0.119.0 (@dehaansa):

    • otelcol.processor.resourcedetection: additional configuration for the ec2 detector to configure retry behavior
    • otelcol.processor.resourcedetection: additional configuration for the gcp detector to collect Managed Instance Group attributes
    • otelcol.processor.resourcedetection: additional configuration for the eks detector to collect cloud account attributes
    • otelcol.processor.resourcedetection: add kubeadm detector to collect local cluster attributes
    • otelcol.processor.cumulativetodelta: add metric_types filtering options
    • otelcol.exporter.awss3: support configuring sending_queue behavior
    • otelcol.exporter.otlphttp: support configuring compression_params, which currently only includes level
    • configtls: opentelemetry components with tls config now support specifying TLS curve preferences
    • sending_queue: opentelemetry exporters with a sending_queue can now configure the queue to be blocking

  • Add go_table_fallback arg to pyroscope.ebpf (@korniltsev)

  • Memory optimizations in pyroscope.scrape (@korniltsev)

  • Do not drop __meta labels in pyroscope.scrape. (@korniltsev)

  • Add the possibility to export span events as logs in otelcol.connector.spanlogs. (@steve-hb)

  • Add json format support for log export via faro receiver (@ravishankar15)

  • (Experimental) Various changes to the experimental component database_observability.mysql:

    • connection_info: add namespace to the metric (@cristiangreco)
    • query_sample: better support for table name parsing (@cristiangreco)
    • query_sample: capture schema name for query samples (@cristiangreco)
    • query_sample: fix error handling during result set iteration (@cristiangreco)
    • query_sample: improve parsing of truncated queries (@cristiangreco)
    • query_sample: split out sql parsing logic to a separate file (@cristiangreco)
    • schema_table: add table columns parsing (@cristiagreco)
    • schema_table: correctly quote schema and table name in SHOW CREATE (@cristiangreco)
    • schema_table: fix handling of view table types when detecting schema (@matthewnolf)
    • schema_table: refactor cache config in schema_table collector (@cristiangreco)
    • Component: add enable/disable collector configurability to database_observability.mysql. This removes the query_samples_enabled argument, now configurable via enable/disable collector. (@fridgepoet)
    • Component: always log instance label key (@cristiangreco)
    • Component: better error handling for collectors (@cristiangreco)
    • Component: use labels for some indexed logs elements (@cristiangreco)

  • Reduce CPU usage of loki.source.windowsevent by up to 85% by updating the bookmark file every 10 seconds instead of after every event and by optimizing the retrieval of the process name. (@wildum)

  • Ensure consistent service_name label handling in pyroscope.receive_http to match Pyroscope's behavior. (@marcsanmi)

  • Improved memory and CPU performance of Prometheus pipelines by changing the underlying implementation of targets (@thampiotr)

  • Add config_merge_strategy in prometheus.exporter.snmp to optionally merge custom snmp config with embedded config instead of replacing. Useful for providing SNMP auths. (@v-zhuravlev)

  • Upgrade beyla.ebpf to v2.0.4. The full list of changes can be found in the Beyla release notes. (@marctc)

Bugfixes

  • Fix log rotation for Windows in loki.source.file by refactoring the component to use the runner pkg. This should also reduce CPU consumption when tailing a lot of files in a dynamic environment. (@wildum)

  • Add livedebugging support for prometheus.remote_write (@ravishankar15)

  • Add livedebugging support for otelcol.connector.* components (@wildum)

  • Bump snmp_exporter and embedded modules to 0.27.0. Add support for multi-module handling by comma separation and expose argument to increase SNMP polling concurrency for prometheus.exporter.snmp. (@v-zhuravlev)

  • Add support for pushv1.PusherService Connect API in pyroscope.receive_http. (@simonswine)

  • Fixed an issue where loki.process would sometimes output live debugging entries out-of-order (@thampiotr)

  • Fixed a bug where components could be evaluated concurrently without the full context during a config reload (@wildum)

  • Fixed locks that wouldn't be released in the remotecfg service if some errors occurred during the configuration reload (@spartan0x117)

  • Fix issue with prometheus.write.queue that lead to excessive connections. (@mattdurham)

  • Fixed a bug where loki.source.awsfirehose and loki.source.gcplog could not be used from within a module. (@tpaschalis)

  • Fix an issue where Prometheus metric name validation scheme was set by default to UTF-8. It is now set back to the previous "legacy" scheme. An experimental flag --feature.prometheus.metric-validation-scheme can be used to switch it to utf-8 to experiment with UTF-8 support. (@thampiotr)

Other changes

  • Upgrading to Prometheus v2.54.1. (@ptodev)

    • discovery.docker has a new match_first_network attribute for matching the first network if the container has multiple networks defined, thus avoiding collecting duplicate targets.

    • discovery.ec2, discovery.kubernetes, discovery.openstack, and discovery.ovhcloud add extra __meta_ labels.

    • prometheus.remote_write supports Azure OAuth and Azure SDK authentication.

    • discovery.linode has a new region attribute, as well as extra __meta_ labels.

    • A new scrape_native_histograms argument for prometheus.scrape. This is enabled by default and can be used to explicitly disable native histogram support. In previous versions of Alloy, native histogram support has also been enabled by default as long as scrape_protocols starts with PrometheusProto.

    • Change the stability of the remotecfg feature from "public preview" to "generally available". (@erikbaranowski)

v1.6.1

Bugs

  • Resolve issue with Beyla starting. (@rafaelroquetto)

v1.6.0

Breaking changes

  • Upgrade to OpenTelemetry Collector v0.116.0:
    • otelcol.processor.tailsampling: Change decision precedence when using and_sub_policy and invert_match. For more information, see the release notes for Alloy 1.6.

Features

  • Add support for TLS to prometheus.write.queue. (@mattdurham)

  • Add otelcol.receiver.syslog component to receive otel logs in syslog format (@dehaansa)

  • Add support for metrics in otelcol.exporter.loadbalancing (@madaraszg-tulip)

  • Add add_cloudwatch_timestamp to prometheus.exporter.cloudwatch metrics. (@captncraig)

  • Add support to prometheus.operator.servicemonitors to allow endpointslice role. (@yoyosir)

  • Add otelcol.exporter.splunkhec allowing to export otel data to Splunk HEC (@adlotsof)

  • Add otelcol.receiver.solace component to receive traces from a Solace broker. (@wildum)

  • Add otelcol.exporter.syslog component to export logs in syslog format (@dehaansa)

  • (Experimental) Add a database_observability.mysql component to collect mysql performance data. (@cristiangreco & @matthewnolf)

  • Add otelcol.receiver.influxdb to convert influx metric into OTEL. (@EHSchmitt4395)

  • Add a new /-/healthy endpoint which returns HTTP 500 if one or more components are unhealthy. (@ptodev)

Enhancements

  • Improved performance by reducing allocation in Prometheus write pipelines by ~30% (@thampiotr)

  • Update prometheus.write.queue to support v2 for cpu performance. (@mattdurham)

  • (Experimental) Add health reporting to database_observability.mysql component (@cristiangreco)

  • Add second metrics sample to the support bundle to provide delta information (@dehaansa)

  • Add all raw configuration files & a copy of the latest remote config to the support bundle (@dehaansa)

  • Add relevant golang environment variables to the support bundle (@dehaansa)

  • Add support for server authentication to otelcol components. (@aidaleuc)

  • Update mysqld_exporter from v0.15.0 to v0.16.0 (including 2ef168bf6), most notable changes: (@cristiangreco)

    • Support MySQL 8.4 replicas syntax
    • Fetch lock time and cpu time from performance schema
    • Fix fetching tmpTables vs tmpDiskTables from performance_schema
    • Skip SPACE_TYPE column for MariaDB >=10.5
    • Fixed parsing of timestamps with non-zero padded days
    • Fix auto_increment metric collection errors caused by using collation in INFORMATION_SCHEMA searches
    • Change processlist query to support ONLY_FULL_GROUP_BY sql_mode
    • Add perf_schema quantile columns to collector

  • Live Debugging button should appear in UI only for supported components (@ravishankar15)

  • Add three new stdlib functions to_base64, from_URLbase64 and to_URLbase64 (@ravishankar15)

  • Add ignore_older_than option for local.file_match (@ravishankar15)

  • Add livedebugging support for discovery components (@ravishankar15)

  • Add livedebugging support for discover.relabel (@ravishankar15)

  • Performance optimization for live debugging feature (@ravishankar15)

  • Upgrade github.com/goccy/go-json to v0.10.4, which reduces the memory consumption of an Alloy instance by 20MB. If Alloy is running certain otelcol components, this reduction will not apply. (@ptodev)

  • improve performance in regexp component: call fmt only if debug is enabled (@r0ka)

  • Update prometheus.write.queue library for performance increases in cpu. (@mattdurham)

  • Update loki.secretfilter to be compatible with the new [[rules.allowlists]] gitleaks allowlist format (@romain-gaillard)

  • Update async-profiler binaries for pyroscope.java to 3.0-fa937db (@aleks-p)

  • Reduced memory allocation in discovery components by up to 30% (@thampiotr)

Bugfixes

  • Fix issue where alloy_prometheus_relabel_metrics_processed was not being incremented. (@mattdurham)

  • Fixed issue with automemlimit logging bad messages and trying to access cgroup on non-linux builds (@dehaansa)

  • Fixed issue with reloading configuration and prometheus metrics duplication in prometheus.write.queue. (@mattdurham)

  • Updated prometheus.write.queue to fix issue with TTL comparing different scales of time. (@mattdurham)

  • Fixed an issue in the prometheus.operator.servicemonitors, prometheus.operator.podmonitors and prometheus.operator.probes to support capitalized actions. (@QuentinBisson)

  • Fixed an issue where the otelcol.processor.interval could not be used because the debug metrics were not set to default. (@wildum)

  • Fixed an issue where loki.secretfilter would crash if the secret was shorter than the partial_mask value. (@romain-gaillard)

  • Change the log level in the eventlogmessage stage of the loki.process component from warn to debug. (@wildum)

  • Fix a bug in loki.source.kafka where the topics argument incorrectly used regex matching instead of exact matches. (@wildum)

Other changes

v1.5.1

Enhancements

  • Logs from underlying clustering library memberlist are now surfaced with correct level (@thampiotr)

  • Allow setting informer_sync_timeout in prometheus.operator.* components. (@captncraig)

  • For sharding targets during clustering, loki.source.podlogs now only takes into account some labels. (@ptodev)

  • Improve instrumentation of pyroscope.relabel component. (@marcsanmi)

Bugfixes

  • Fixed an issue in the pyroscope.write component to prevent TLS connection churn to Pyroscope when the pyroscope.receive_http clients don't request keepalive (@madaraszg-tulip)

  • Fixed an issue in the pyroscope.write component with multiple endpoints not working correctly for forwarding profiles from pyroscope.receive_http (@madaraszg-tulip)

  • Fixed a few race conditions that could lead to a deadlock when using import statements, which could lead to a memory leak on /metrics endpoint of an Alloy instance. (@thampiotr)

  • Fix a race condition where the ui service was dependent on starting after the remotecfg service, which is not guaranteed. (@dehaansa & @erikbaranowski)

  • Fixed an issue in the otelcol.exporter.prometheus component that would set series value incorrectly for stale metrics (@YusifAghalar)

  • loki.source.podlogs: Fixed a bug which prevented clustering from working and caused duplicate logs to be sent. The bug only happened when no selector or namespace_selector blocks were specified in the Alloy configuration. (@ptodev)

  • Fixed an issue in the pyroscope.write component to allow slashes in application names in the same way it is done in the Pyroscope push API (@marcsanmi)

  • Fixed a crash when updating the configuration of remote.http. (@kinolaev)

  • Fixed an issue in the otelcol.processor.attribute component where the actions delete and hash could not be used with the pattern argument. (@wildum)

  • Fixed an issue in the prometheus.exporter.postgres component that would leak goroutines when the target was not reachable (@dehaansa)

v1.5.0

Breaking changes

Features

  • Add support bundle generation via the API endpoint /-/support (@dehaansa)

  • Add the function path_join to the stdlib. (@wildum)

  • Add pyroscope.receive_http component to receive and forward Pyroscope profiles (@marcsanmi)

  • Add support to loki.source.syslog for the RFC3164 format ("BSD syslog"). (@sushain97)

  • Add support to loki.source.api to be able to extract the tenant from the HTTP X-Scope-OrgID header (@QuentinBisson)

  • (Experimental) Add a loki.secretfilter component to redact secrets from collected logs.

  • (Experimental) Add a prometheus.write.queue component to add an alternative to prometheus.remote_write which allowing the writing of metrics to a prometheus endpoint. (@mattdurham)

  • (Experimental) Add the array.combine_maps function to the stdlib. (@ptodev, @wildum)

Enhancements

  • The mimir.rules.kubernetes component now supports adding extra label matchers to all queries discovered via PrometheusRule CRDs. (@thampiotr)

  • The cluster.use-discovery-v1 flag is now deprecated since there were no issues found with the v2 cluster discovery mechanism. (@thampiotr)

  • SNMP exporter now supports labels in both target and targets parameters. (@mattdurham)

  • Add support for relative paths to import.file. This new functionality allows users to use import.file blocks in modules imported via import.git and other import.file. (@wildum)

  • prometheus.exporter.cloudwatch: The discovery block now has a recently_active_only configuration attribute to return only metrics which have been active in the last 3 hours.

  • Add Prometheus bearer authentication to a prometheus.write.queue component (@freak12techno)

  • Support logs that have a timestamp field instead of a time field for the loki.source.azure_event_hubs component. (@andriikushch)

  • Add proxy_url to otelcol.exporter.otlphttp. (@wildum)

  • Allow setting informer_sync_timeout in prometheus.operator.* components. (@captncraig)

Bugfixes

  • Fixed a bug in import.git which caused a "non-fast-forward update" error message. (@ptodev)

  • Do not log error on clean shutdown of loki.source.journal. (@thampiotr)

  • prometheus.operator.* components: Fixed a bug which would sometimes cause a "failed to create service discovery refresh metrics" error after a config reload. (@ptodev)

Other changes

v1.4.3

Bugfixes

  • Fix an issue where some faro.receiver would drop multiple fields defined in payload.meta.browser, as fields were defined in the struct.

  • pyroscope.scrape no longer tries to scrape endpoints which are not active targets anymore. (@wildum @mattdurham @dehaansa @ptodev)

  • Fixed a bug with loki.source.podlogs not starting in large clusters due to short informer sync timeout. (@elburnetto-intapp)

  • prometheus.exporter.windows: Fixed bug with exclude regular expression config arguments which caused missing metrics. (@ptodev)

v1.4.2

Bugfixes

  • Update windows_exporter from v0.27.2 vo v0.27.3: (@jkroepke)

    • Fixes a bug where scraping Windows service crashes alloy

  • Update yet-another-cloudwatch-exporter from v0.60.0 vo v0.61.0: (@morremeyer)

    • Fixes a bug where cloudwatch S3 metrics are reported as 0

  • Issue 1687 - otelcol.exporter.awss3 fails to configure (@cydergoth)

    • Fix parsing of the Level configuration attribute in debug_metrics config block
    • Ensure "optional" debug_metrics config block really is optional

  • Fixed an issue with loki.process where stage.luhn and stage.timestamp would not apply default configuration settings correctly (@thampiotr)

  • Fixed an issue with loki.process where configuration could be reloaded even if there were no changes. (@ptodev, @thampiotr)

  • Fix issue where loki.source.kubernetes took into account all labels, instead of specific logs labels. Resulting in duplication. (@mattdurham)

v1.4.1

Bugfixes

  • Windows installer: Don't quote Alloy's binary path in the Windows Registry. (@jkroepke)

v1.4.0

Security fixes

  • Add quotes to windows service path to prevent path interception attack. CVE-2024-8975 (@mattdurham)

Breaking changes

  • Some debug metrics for otelcol components have changed. (@thampiotr) For example, otelcol.exporter.otlp's exporter_sent_spans_ratio_total metric is now otelcol_exporter_sent_spans_total.

  • [otelcol.processor.transform] The functions convert_sum_to_gauge and convert_gauge_to_sum must now be used in the metric context rather than in the datapoint context. open-telemetry/opentelemetry-collector-contrib#34567 (@wildum)

  • Upgrade Beyla from 1.7.0 to 1.8.2. A complete list of changes can be found on the Beyla releases page: https://github.com/grafana/beyla/releases. (@wildum) It contains a few breaking changes for the component beyla.ebpf:

    • renamed metric process.cpu.state to cpu.mode
    • renamed metric beyla_build_info to beyla_internal_build_info

Features

  • Added Datadog Exporter community component, enabling exporting of otel-formatted Metrics and traces to Datadog. (@polyrain)
  • (Experimental) Add an otelcol.processor.interval component to aggregate metrics and periodically forward the latest values to the next component in the pipeline.

Enhancements

  • Clustering peer resolution through --cluster.join-addresses flag has been improved with more consistent behaviour, better error handling and added support for A/AAAA DNS records. If necessary, users can temporarily opt out of this new behaviour with the --cluster.use-discovery-v1, but this can only be used as a temporary measure, since this flag will be disabled in future releases. (@thampiotr)

  • Added a new panel to Cluster Overview dashboard to show the number of peers seen by each instance in the cluster. This can help diagnose cluster split brain issues. (@thampiotr)

  • Updated Snowflake exporter with performance improvements for larger environments. Also added a new panel to track deleted tables to the Snowflake mixin. (@Caleb-Hurshman)

  • Add a otelcol.processor.groupbyattrs component to reassociate collected metrics that match specified attributes from opentelemetry. (@kehindesalaam)

  • Update windows_exporter to v0.27.2. (@jkroepke) The smb.enabled_list and smb_client.enabled_list doesn't have any effect anymore. All sub-collectors are enabled by default.

  • Live debugging of loki.process will now also print the timestamp of incoming and outgoing log lines. This is helpful for debugging stage.timestamp. (@ptodev)

  • Add extra validation in beyla.ebpf to avoid panics when network feature is enabled. (@marctc)

  • A new parameter aws_sdk_version_v2 is added for the cloudwatch exporters configuration. It enables the use of aws sdk v2 which has shown to have significant performance benefits. (@kgeckhart, @andriikushch)

  • prometheus.exporter.cloudwatch can now collect metrics from custom namespaces via the custom_namespace block. (@ptodev)

  • Add the label alloy_cluster in the metric alloy_config_hash when the flag cluster.name is set to help differentiate between configs from the same alloy cluster or different alloy clusters. (@wildum)

  • Add support for discovering the cgroup path(s) of a process in process.discovery. (@mahendrapaipuri)

Bugfixes

  • Fix a bug where the scrape timeout for a Probe resource was not applied, overwriting the scrape interval instead. (@morremeyer, @stefanandres)

  • Fix a bug where custom components don't always get updated when the config is modified in an imported directory. (@ante012)

  • Fixed an issue which caused loss of context data in Faro exception. (@codecapitano)

  • Fixed an issue where providing multiple hostnames or IP addresses via --cluster.join-addresses would only use the first provided value. (@thampiotr)

  • Fixed an issue where providing <hostname>:<port> in --cluster.join-addresses would only resolve with DNS to a single address, instead of using all the available records. (@thampiotr)

  • Fixed an issue where clustering peers resolution via hostname in --cluster.join-addresses resolves to duplicated IP addresses when using SRV records. (@thampiotr)

  • Fixed an issue where the connection_string for the loki.source.azure_event_hubs component was displayed in the UI in plaintext. (@MorrisWitthein)

  • Fix a bug in discovery.* components where old targets would continue to be exported to downstream components. This would only happen if the config for discovery.* is reloaded in such a way that no new targets were discovered. (@ptodev, @thampiotr)

  • Fixed bug in loki.process with sampling stage where all components use same drop_counter_reason. (@captncraig)

  • Fixed an issue (see #1599) where specifying both path and key in the remote.vault path configuration could result in incorrect URLs. The path and key arguments have been separated to allow for clear and accurate specification of Vault secrets. (@PatMis16)

Other

v1.3.4

Bugfixes

  • Windows installer: Don't quote Alloy's binary path in the Windows Registry. (@jkroepke)

v1.3.2

Security fixes

  • Add quotes to windows service path to prevent path interception attack. CVE-2024-8975 (@mattdurham)

v1.3.1

Bugfixes

  • Changed the cluster startup behaviour, reverting to the previous logic where a failure to resolve cluster join peers results in the node creating its own cluster. This is to facilitate the process of bootstrapping a new cluster following user feedback (@thampiotr)

  • Fix a memory leak which would occur any time loki.process had its configuration reloaded. (@ptodev)

v1.3.0

Breaking changes

  • [otelcol.exporter.otlp,otelcol.exporter.loadbalancing]: Change the default gRPC load balancing strategy. The default value for the balancer_name attribute has changed to round_robin open-telemetry/opentelemetry-collector#10319

Breaking changes to non-GA functionality

  • Update Public preview remotecfg argument from metadata to attributes. (@erikbaranowski)

  • The default value of the argument unmatched in the block routes of the component beyla.ebpf was changed from unset to heuristic (@marctc)

Features

  • Added community components support, enabling community members to implement and maintain components. (@wildum)

  • A new otelcol.exporter.debug component for printing OTel telemetry from other otelcol components to the console. (@BarunKGP)

Enhancements

Bugfixes

  • Fixed a clustering mode issue where a fatal startup failure of the clustering service would exit the service silently, without also exiting the Alloy process. (@thampiotr)

  • Fix a bug which prevented config reloads to work if a Loki metrics stage is in the pipeline. Previously, the reload would fail for loki.process without an error in the logs and the metrics from the metrics stage would get stuck at the same values. (@ptodev)

v1.2.1

Bugfixes

  • Fixed an issue with loki.source.kubernetes_events not starting in large clusters due to short informer sync timeout. (@nrwiersma)

  • Updated ckit to fix an issue with armv7 panic on startup when forming a cluster. (@imavroukakis)

  • Fixed a clustering mode issue where a failure to perform static peers discovery did not result in a fatal failure at startup and could lead to potential split-brain issues. (@thampiotr)

Other

  • Use Go 1.22.5 for builds. (@mattdurham)

v1.2.0

Security fixes

Breaking changes

  • Updated OpenTelemetry to v0.102.1. (@mattdurham)
    • Components otelcol.receiver.otlp,otelcol.receiver.zipkin,otelcol.extension.jaeger_remote_sampling, and otelcol.receiver.jaeger setting max_request_body_size default changed from unlimited size to 20MiB. This is due to CVE-2024-36129.

Breaking changes to non-GA functionality

  • Update Public preview remotecfg to use alloy-remote-config instead of agent-remote-config. The API has been updated to use the term collector over agent. (@erikbaranowski)

  • Component otelcol.receiver.vcenter removed vcenter.host.network.packet.errors, vcenter.host.network.packet.count, and vcenter.vm.network.packet.count.

    • vcenter.host.network.packet.errors replaced by vcenter.host.network.packet.error.rate.
    • vcenter.host.network.packet.count replaced by vcenter.host.network.packet.rate.
    • vcenter.vm.network.packet.count replaced by vcenter.vm.network.packet.rate.

Features

  • Add an otelcol.exporter.kafka component to send OTLP metrics, logs, and traces to Kafka.

  • Added live debugging to the UI. Live debugging streams data as they flow through components for debugging telemetry data. Individual components must be updated to support live debugging. (@wildum)

  • Added live debugging support for prometheus.relabel. (@wildum)

  • (Experimental) Add a otelcol.processor.deltatocumulative component to convert metrics from delta temporality to cumulative by accumulating samples in memory. (@rfratto)

  • (Experimental) Add an otelcol.receiver.datadog component to receive metrics and traces from Datadog. (@carrieedwards, @jesusvazquez, @alexgreenbank, @fedetorres93)

  • Add a prometheus.exporter.catchpoint component to collect metrics from Catchpoint. (@bominrahmani)

  • Add the -t/--test flag to alloy fmt to check if a alloy config file is formatted correctly. (@kavfixnel)

Enhancements

  • (Public preview) Add native histogram support to otelcol.receiver.prometheus. (@wildum)

  • (Public preview) Add metrics to report status of remotecfg service. (@captncraig)

  • Added scrape_protocols option to prometheus.scrape, which allows to control the preferred order of scrape protocols. (@thampiotr)

  • Add support for configuring CPU profile's duration scraped by pyroscope.scrape. (@hainenber)

  • prometheus.exporter.snowflake: Add support for RSA key-pair authentication. (@Caleb-Hurshman)

  • Improved filesystem error handling when working with loki.source.file and local.file_match, which removes some false-positive error log messages on Windows (@thampiotr)

  • Updates processor/probabilistic_sampler to use new FailedClosed field from OTEL release v0.101.0. (@StefanKurek)

  • Updates receiver/vcenter to use new features and bugfixes introduced in OTEL releases v0.100.0 and v0.101.0. Refer to the v0.100.0 and v0.101.0 release notes for more detailed information. Changes that directly affected the configuration are as follows: (@StefanKurek)

    • The resource attribute vcenter.datacenter.name has been added and enabled by default for all resource types.
    • The resource attribute vcenter.virtual_app.inventory_path has been added and enabled by default to differentiate between resource pools and virtual apps.
    • The resource attribute vcenter.virtual_app.name has been added and enabled by default to differentiate between resource pools and virtual apps.
    • The resource attribute vcenter.vm_template.id has been added and enabled by default to differentiate between virtual machines and virtual machine templates.
    • The resource attribute vcenter.vm_template.name has been added and enabled by default to differentiate between virtual machines and virtual machine templates.
    • The metric vcenter.cluster.memory.used has been removed.
    • The metric vcenter.vm.network.packet.drop.rate has been added and enabled by default.
    • The metric vcenter.cluster.vm_template.count has been added and enabled by default.

  • Add yaml_decode to standard library. (@mattdurham, @djcode)

  • Allow override debug metrics level for otelcol.* components. (@hainenber)

  • Add an initial lower limit of 10 seconds for the the poll_frequency argument in the remotecfg block. (@tpaschalis)

  • Add a constant jitter to remotecfg service's polling. (@tpaschalis)

  • Added support for NS records to discovery.dns. (@djcode)

  • Improved clustering use cases for tracking GCP delta metrics in the prometheus.exporter.gcp (@kgeckhart)

  • Add the targets argument to the prometheus.exporter.snmp component to support passing SNMP targets at runtime. (@wildum)

  • Prefix Faro measurement values with value_ to align with the latest Faro cloud receiver updates. (@codecapitano)

  • Add base64_decode to standard library. (@hainenber)

  • Updated OpenTelemetry Contrib to v0.102.0. (@mattdurham)

    • otelcol.processor.resourcedetection: Added a tags config argument to the azure detection mechanism. It exposes regex-matched Azure resource tags as OpenTelemetry resource attributes.

  • A new snmp_context configuration argument for prometheus.exporter.snmp which overrides the context_name parameter in the SNMP configuration file. (@ptodev)

  • Add extra configuration options for beyla.ebpf to select Kubernetes objects to monitor. (@marctc)

Bugfixes

  • Fixed an issue with prometheus.scrape in which targets that move from one cluster instance to another could have a staleness marker inserted and result in a gap in metrics (@thampiotr)

  • Fix panic when import.git is given a revision that does not exist on the remote repo. (@hainenber)

  • Fixed an issue with loki.source.docker where collecting logs from targets configured with multiple networks would result in errors. (@wildum)

  • Fixed an issue where converting OpenTelemetry Collector configs with unused telemetry types resulted in those types being explicitly configured with an empty array in output blocks, rather than them being omitted entirely. (@rfratto)

Other changes

  • pyroscope.ebpf, pyroscope.java, pyroscope.scrape, pyroscope.write and discovery.process components are now GA. (@korniltsev)

  • prometheus.exporter.snmp: Updating SNMP exporter from v0.24.1 to v0.26.0. (@ptodev, @erikbaranowski)

  • prometheus.scrape component's enable_protobuf_negotiation argument is now deprecated and will be removed in a future major release. Use scrape_protocols instead and refer to prometheus.scrape reference documentation for further details. (@thampiotr)

  • Updated Prometheus dependency to v2.51.2 (@thampiotr)

  • Upgrade Beyla from v1.5.1 to v1.6.3. (@marctc)

v1.1.1

Bugfixes

  • Fix panic when component ID contains / in otelcomponent.MustNewType(ID).(@qclaogui)

  • Exit Alloy immediately if the port it runs on is not available. This port can be configured with --server.http.listen-addr or using the default listen address127.0.0.1:12345. (@mattdurham)

  • Fix a panic in loki.source.docker when trying to stop a target that was never started. (@wildum)

  • Fix error on boot when using IPv6 advertise addresses without explicitly specifying a port. (@matthewpi)

  • Fix an issue where having long component labels (>63 chars) on otelcol.auth components lead to a panic. (@tpaschalis)

  • Update prometheus.exporter.snowflake with the latest version of the exporter as of May 28, 2024 (@StefanKurek)

    • Fixes issue where returned NULL values from database could cause unexpected errors.

  • Bubble up SSH key conversion error to facilitate failed import.git. (@hainenber)

v1.1.0

Features

  • (Public preview) Add support for setting GOMEMLIMIT based on cgroup setting. (@mattdurham)

  • (Experimental) A new otelcol.exporter.awss3 component for sending telemetry data to a S3 bucket. (@Imshelledin21)

  • (Public preview) Introduce BoringCrypto Docker images. The BoringCrypto image is tagged with the -boringcrypto suffix and is only available on AMD64 and ARM64 Linux containers. (@rfratto, @mattdurham)

  • (Public preview) Introduce boringcrypto release assets. BoringCrypto builds are publshed for Linux on AMD64 and ARM64 platforms. (@rfratto, @mattdurham)

  • otelcol.exporter.loadbalancing: Add a new aws_cloud_map resolver. (@ptodev)

  • Introduce a otelcol.receiver.file_stats component from the upstream OpenTelemetry filestatsreceiver component. (@rfratto)

Enhancements

  • Update prometheus.exporter.kafka with the following functionalities (@wildum):

    • GSSAPI config
    • enable/disable PA_FX_FAST
    • set a TLS server name
    • show the offset/lag for all consumer group or only the connected ones
    • set the minimum number of topics to monitor
    • enable/disable auto-creation of requested topics if they don't already exist
    • regex to exclude topics / groups
    • added metric kafka_broker_info

  • In prometheus.exporter.kafka, the interpolation table used to compute estimated lag metrics is now pruned on metadata_refresh_interval instead of prune_interval_seconds. (@wildum)

  • Don't restart tailers in loki.source.kubernetes component by above-average time deltas if K8s version is >= 1.29.1 (@hainenber)

  • In mimir.rules.kubernetes, add support for running in a cluster of Alloy instances by electing a single instance as the leader for the mimir.rules.kubernetes component to avoid conflicts when making calls to the Mimir API. (@56quarters)

  • Add the possibility of setting custom labels for the AWS Firehose logs via X-Amz-Firehose-Common-Attributes header. (@andriikushch)

Bugfixes

  • Fixed issue with defaults for Beyla component not being applied correctly. (marctc)

  • Fix an issue on Windows where uninstalling Alloy did not remove it from the Add/Remove programs list. (@rfratto)

  • Fixed issue where text labels displayed outside of component node's boundary. (@hainenber)

  • Fix a bug where a topic was claimed by the wrong consumer type in otelcol.receiver.kafka. (@wildum)

  • Fix an issue where nested import.git config blocks could conflict if they had the same labels. (@wildum)

  • In mimir.rules.kubernetes, fix an issue where unrecoverable errors from the Mimir API were retried. (@56quarters)

  • Fix an issue where faro.receiver's extra_log_labels with empty value don't map existing value in log line. (@hainenber)

  • Fix an issue where prometheus.remote_write only queued data for sending every 15 seconds instead of as soon as data was written to the WAL. (@rfratto)

  • Imported code using slog logging will now not panic and replay correctly when logged before the logging config block is initialized. (@mattdurham)

  • Fix a bug where custom components would not shadow the stdlib. If you have a module whose name conflicts with an stdlib function and if you use this exact function in your config, then you will need to rename your module. (@wildum)

  • Fix an issue where loki.source.docker stops collecting logs after a container restart. (@wildum)

  • Upgrading pyroscope/ebpf from 0.4.6 to 0.4.7 (@korniltsev):

    • detect libc version properly when libc file name is libc-2.31.so and not libc.so.6
    • treat elf files with short build id (8 bytes) properly

Other changes

v1.0.0

Features

  • Support for programmable pipelines using a rich expression-based syntax.

  • Over 130 components for processing, transforming, and exporting telemetry data.

  • Native support for Kubernetes and Prometheus Operator without needing to deploy or learn a separate Kubernetes operator.

  • Support for creating and sharing custom components.

  • Support for forming a cluster of Alloy instances for automatic workload distribution.

  • (Public preview) Support for receiving configuration from a server for centralized configuration management.

  • A built-in UI for visualizing and debugging pipelines.