v4.3.1

Released 2023-12-14

Fixed

• fix(metrics): use targetallocator serviceaccount created by the operator #3447

v4.3.0

Released 2023-12-13

Added

• feat: add description including helm version to collector #3423
• feat(chart): add global nodeSelector option #3427

Changed

• chore: update OpenTelemetry Collector to v0.90.1-sumo-0 #3438
• chore: update OpenTelemetry Operator to v0.44.0 #3441

Fixed

• fix(metrics): use sumologic.metrics.excludeNamespaceRegex instead of sumologic.logs.container.excludeNamespaceRegex #3428
• fix: fix add_timestamp behavior #3434

v4.2.0

Released 2023-11-27

Breaking Changes

• chore!: upgrade otel to 0.89.0-sumo-2 #3368 #3414

  Otelcol enqueue failure metrics are now only reported if they're non-zero. If you have monitors set based on the assumption that they're always present, you'll need to update them. Monitors installed by this Chart are unaffected.

Added

• feat(metrics): add setting affinity for metrics collector #3400

Changed

• chore: add support for:
  • EKS 1.28 #3387
  • EKS Fargate 1.28 #3387
  • OpenShift 4.14 #3387
  • kops 1.28 #3387
  • AKS 1.28 #3387
  • GKE v1.28 #3403
• chore: drop support for:
  • EKS 1.23 #3387
  • kops 1.23 #3387
  • OpenShift 4.10 #3387
• deps: update tailing-sidecar to 0.9.0 #3391
• deps: update metrics-server to 6.6.3 #3392, #3405, #3416
• deps: update falco to v3.8.5 #3393, #3415
• deps: update opentelemetry-operator to v0.43.0 #3394, #3404, #3418
• deps: upgrade otel to 0.89.0-sumo-2 #3414
• deps: update telegraf operator to v1.3.12 #3417

Fixed

• fix(metrics): set nodeSelector and tolerations for target allocator #3411
• fix(metrics): set affinity for target allocator #3412

v4.1.0

Released 2023-11-03

Added

• feat: allow setting the cluster DNS domain #3362
• Add new service account for the otel cloudwatch collector statefulset #3374

Changed

• chore: upgrade nginx image to 1.25.2-alpine-sumo-1 #3375

Fixed

• fix(otel-collector): deploy collector by default on all nodes #3348
• use autoscaling/v2 if available on the cluster #3366
• fix(instrumentation): replace tools image with kubectl #3373

v3.17.0

Released 2023-11-03

Added

• Add new service account for the otel cloudwatch collector statefulset #3374

Changed

• chore: drop support for GKE with k8s 1.23 #3342
• chore: upgrade nginx image to 1.25.2-alpine-sumo-1 #3375

Fixed

• fix(otel-collector): deploy collector by default on all nodes #3348
• use autoscaling/v2 if available on the cluster #3367
• fix(instrumentation): replace tools image with kubectl #3373

v4.0.1

Released 2023-10-25

This release mitigates open-telemetry/opentelemetry-collector-contrib#28567 by downgrading the otel version.

Fixed

• fix: downgrade otel to 0.86.0-sumo-1 #3352

v3.16.2

Released 2023-10-25

This release mitigates open-telemetry/opentelemetry-collector-contrib#28567 by downgrading the otel version.

Fixed

• fix: downgrade otel to 0.86.0-sumo-1 #3352

v4.0.0

Released 2023-10-20

This release replaces Prometheus with the OpenTelemetry Collector by default, thus completing our transition to Otel. A fresh installation of the Chart will use the OpenTelemetry Collector for all data collection and processing. In addition, data is now sent to Sumo using the Otel-native OTLP protocol.

Prometheus is still available as a non-default option. However, Fluentd and Fluent Bit have been completely removed, so installations using them will need migrate to Otel for metadata before being able to upgrade to v4.

We've done our best to make this transition as painless as possible. The majority of users should be able to upgrade without any changes to the configuration. Please check out the migration guide linked below for details.

Migration from v3

See the migration guide for details.

Breaking Changes

• feat!: remove support for fluent-bit and fluentd #3244
• feat!: truncate fullname after 22 characters #3248
• feat(metrics)!: use otel by default #3284
• feat!: use OTLP sources by default #3297
• feat!(metrics): move extra processors after sumologic_schema #3306
• fix(metrics)!: drop k8s.node.name attribute #3295
• feat!: enable autoscaling by default #3329

Added

• feat(logs): add sumologic.logs.additionalFields property #3286
• feat(metrics): add additionalServiceMonitors setting #3292
• feat(metrics): collect node_memory_MemAvailable_bytes #3322
• chore: add support for k8s 1.27 with KOPS #3332

Changed

• feat(prometheus): Removing prometheus recording rules #3211
• feat(metrics): move app metrics filtering to metadata layer #3232
• chore: drop support for GKE with k8s 1.23 #3340

Fixed

• fix(openshift): fix SecurityContextConstraints #3308 #3309 #3310

v3.16.1

Released 2023-10-20

Changed

• chore: upgrade otel to 0.87.0-sumo-0 #3334

v3.16.0

Released 2023-10-18

Added

• feat(logs): add sumologic.logs.additionalFields property #3286
• feat(metrics): add additionalServiceMonitors setting #3292
• feat(metrics): allow overriding metrics collector configuration #3314

Changed

• chore: update setup job to v3.11.0 #3320

Fixed

• fix(logs)!: move JSON parsing after user-defined processors #3281

  The log body will now always be a string if accessed in extra processors. Users who want to access specific fields in their parsed JSON
  log should explicitly call ParseJSON in their processor definition.

• fix(metrics): decompose OTLP histograms #3289

• fix(metrics): drop stale datapoints #3318

• fix: fix Otel Operator installation with Helm 3.13 #3321

• fix(metrics): kube-state-metrics pod metadata #3323