Insight Release Notes¶

This page lists the Release Notes of Insight, so that you can understand the evolution path and feature changes of each version.

2024.03.31¶

v0.25.0¶

These changes are for Insight Server.

Improvements¶

• Improved Log correlation support for filtering based on TraceID and container groups
• Improved Support for encrypting sensitive information of notification objects
• Improved Split insight-server into insight-server and insight-manager components
• Improved High availability support for opentelemetry-collector component
• Improved Grafana and Jaeger are not accessible without login
• Improved Support for customizing whether to initialize log indexes during installation

Bug Fixes¶

• Fixed permission issues with alarm-related APIs
• Fixed issue with no data in overview query metrics
• Fixed validation issue when importing YAML for alarm policies
• Fixed limitation issue with Grafana access speed

2024.01.31¶

Insight Server: v0.24.0¶

New Features¶

• Added Support for alarm suppression
• Added Support for alarm templates and creating alarm policies from templates

Improvements¶

• Improved Grafana supports adding data sources of JSON API type
• Improved Grafana prevents using the ECS key to exit fullscreen mode

Fixes¶

• Fixed Inaccurate preview queries when creating log alarms
• Fixed Error in listening IPv6 during insight-system deployment
• Fixed Issue with Grafana dashboards not being accessible without authentication
• Fixed Enabling InsecureSkipVerify for all SMTP emails

Insight Agent: v0.23.0¶

Improvements¶

• Improved Upgraded image versions of related components for OpenTelemetry Collector

2023.12.31¶

Insight Server: v0.23.0¶

New Features¶

• Added integration with network observation Deepflow Community Edition
• Added alert templates

Improvements¶

• Improved usage of DSL as the query statement for log alerts

Fixes¶

• Fixed an issue that topology graph does not have request latency data
• Fixed an issue where Insight Agent status is not updated when cluster status is abnormal
• Fixed an issue of missing metadata in the configuration file for probing
• Fixed inaccurate query and incorrect alert objects in the preview trend chart of log alerts
• Fixed duplicate occurrence of app.kubernetes.io/name field in Insight Server Chart

Insight Agent: v0.23.0¶

Improvements¶

• Improved automatic injection of k8s_namespce_name for probe injection in distributed tracing
• Improved automatic generation and configuration of log indexes for each deployed Insight Agent cluster

2023.11.30¶

Insight Server: v0.22.0¶

New Features¶

• Added support for network connectivity testing
• Added ability to import alert policies via YAML
• Added support for operation audit logging

Improvements¶

• Improved automatic deployment of OTel instrumentation CR during installation

Fixes¶

• Fixed an issue with Elasticsearch index initialization failure

Insight Agent: v0.22.0¶

• Fixed compatibility issue with Fluentbit directory collection and DCE 4.0

2023.10.31¶

Insight Server: v0.21.0¶

New Features¶

• Added Namespace-level monitoring.

Improvements¶

• Improved navigation structure in Insight.
• Improved a behavior of clicking on a link distribution graph allows quick access to proper link details.

Fixes¶

• Fixed an issue of unable to view container log context.
• Fixed an error when initializing event index.

Insight Agent: v0.21.0¶

Bug Fixes¶

• Fixed an issue of abnormal span names appearing in link queries.
• Fixed an issue where containers started by the tailing-sidecar in the Bank Kirin Kylin-V10 (SP3) operating system could not start properly.

2023.08.31¶

Insight Server: v0.20.0¶

New Features¶

• Added container event alerts
• Added linked query-related logs
• Added metadata added to event details
• Added support for Lucene syntax queries in logs

Improvements¶

• Added prompts for cluster status abnormalities
• Improved filtering conditions for logs
• Improved duplicate tags when creating silent alert conditions
• Improved node logs now support filtering by file path
• Improved built-in alerting policies for CoreDNS.

Bug Fixes¶

• Fixed creation time error for Elasticsearch in system components.
• Fixed discrepancy between the number of logs retrieved in contextual log queries and the actual count.
• Fixed redirection issue with repair suggestions in alerts.
• Fixed compatibility issues with certain data in event queries.
• Fixed color error in the legend of the topology graph.
• Fixed an issue where all nodes became external nodes in the topology diagram when canceling cluster and namespace boundaries.
• Fixed failure to send alerts if there is an error in the content of any notification type in the message template.

Insight Agent: v0.20.0¶

Improvements¶

• Improved support for consuming log and linked data via Kafka

Bug Fixes¶

• Fixed compatibility issues with Openshift clusters
• Fixed compatibility issues with Kubernetes v0.18.20
• Fixed compatibility issues with DCE 4.0
• Fixed issue with abnormal metric calculations

2023.07.30¶

v0.19.0¶

New Features¶

• Added kubernetes event analysis and query functionality.
• Added built-in alert policy provides suggestions for alert fixes.
• Added support for customizing log export fields and formats.

Improvements¶

• Improved logic for retrieving Elasticsearch component status in system components.
• Improved support for configuring whether to import default message templates in installation parameters.

Fixes¶

• Fixed a permissions issue for some roles when viewing alert rule details.
• Fixed a permissions issue for some roles when creating, editing, or previewing silent alerts.
• Fixed a permissions issue for some roles when viewing alert message details.
• Fixed time display issue in bubble chart for trace queries.
• Fixed an issue with missing namespace parameter when querying container logs.
• Fixed incorrect custom tag parameter for default variables in message templates.
• Fixed an issue of no new request sent when switching filter conditions.
• Fixed a permissions issue for creating, editing, or previewing silent alerts.
• Fixed an issue with log statistics value being displayed as 0 in overview.
• Fixed an issue with some built-in alert policies not taking effect.

2023.07.01¶

v0.18.0¶

New Features¶

• Added log alerts.
• Added cluster status to collection management.
• Added silence conditions to the silent rule list.
• Added email configuration detection.
• Added support for PostgreSQL and Kingbase as system databases.
• Added Nvidia GPU monitoring dashboard.

Improvements¶

• Updated the legend for updating service topology.
• Added average value of metrics and sorting support to operation metrics in service details.
• Added sorting by span, latency, occurrence time, etc. to link queries.
• Added search functionality to the dropdown menu for notification configuration in alert policies.
• Added timezone formatting support to alert templates.
• Upgraded opentelemetry collector Chart version from 0.50.1 to 0.59.3 .
• Upgraded opentelemetry Operator Chart version from 0.26.1 to 0.30.1 .

Fixes¶

• Fixed an issue where the previewed silent alerts did not match the actual silenced alerts.
• Fixed an issue with incorrect component versions in system components.
• Fixed the error in the title of the list in the dashboard.
• Fixed an issue where no data was returned when previewing matching alerts while creating silence through the alert list.
• Fixed an issue where Fluentbit could not start for the first time in some environments.

2023.06.01¶

v0.17.0¶

Added¶

• Added support for viewing active alert and history alert details.
• Added support for quickly creating silent rules through alerts.
• Added support for SMS notifications in alerts.
• Added support for sending test messages in email notifications.
• Added support for customizing email subjects in message templates for email notifications.
• Added variable explanations to message templates.
• Added default high availability support for Insight Server component.

Improvements¶

• Improved display full TraceID in trace query.
• Improved added prompt when service is empty in trace query.
• Improved added no data prompt for JVM monitoring.
• Improved alert policy details do not show other parameters if not notified.
• Improved added resource limits for OpenTelemetry Operator.
• Improved upgraded Grafana version to v9.3.14.
• Improved upgraded tailing sidecar version from v0.5.6 to v0.7.0.
• Improved upgraded kube-prometheus-stack version to v45.28.1.
• Improved upgraded prometheus version to v2.44.0.

Bug Fixes¶

• Fixed an issue where service topology was effective when namespace was empty.
• Fixed an issue with invalid documentation link in service topology status description.
• Fixed duplication of namespace and stateless workload in displayed namespace when creating alert policies.
• Fixed missing trigger value data in alert list and alert policy in alerts.
• Fixed validation error in required fields for type, keyword, and value when adding silence conditions in alert silencing.
• Fixed time zone not being effective in alert silencing time range.
• Fixed an issue where no alert object was created for promQL rules for workload types.
• Fixed an issue where modifications to built-in alert rules could not be saved.
• Fixed an issue with hardcoded time zone in components.

2023.04.28¶

v0.16.0¶

New Features¶

• Added JVM monitoring for Java applications.
• Added Set alert silent notification.
• Added alert policy function, which supports adding multiple alert rules in an alert policy for management.
• Added nginx Ingress, Contour and other components monitoring dashboard.
• Added The service topology supports enabling or disabling virtual nodes.
• Added statistics on the number of currently active alerts.
• Added ServiceMonitor with built-in HwameiStor component.
• Added domain name supports access to sub path.
• Added [Dashboard] Add record information according to the global management configuration
• Added [Dashboard] domain name supports access to sub path.

Improvements¶

• Improved the content of the built-in message template.
• Improved Adjust the built-in alert rules to the corresponding policies.
• Improved The filter display problem when the service namespace is empty.
• Improved Add cache to service list.

Fixes¶

• Fixed The problem of Chinese invalid search.
• Fixed The problem that the service topology namespace permission cannot view the topology map.
• Fixed [Dashboard] Kylin icon style problem
• Fixed [dashboard] tooltip is too long

2023.04.04¶

v0.15.4¶

Improvements¶

• Improved the default number of primary shards for ES index to match the default number of ES nodes in the middleware.
• Improved the JVM information API response body structure.

Fixes¶

• Fixed multiple trigger alert logging
• Fixed Fluentbit CVE-2021-46848, upgrade from 2.0.5 to 2.0.8
• Fixed task to check license resources
• Fixed SQL statement to clear alert history
• Fixed the Chinese problem in the English dashboard

2023.03.30¶

v0.15.1¶

New Features¶

• Added JVM metric collection and integrated monitoring panel
• Added Link Access Guidance
• Added service topology supports error rate and request delay filtering
• Added link supports Trace ID search
• Added Prometheus component to enable automatic vertical expansion

Improvements¶

• Improved the style of the topology map virtual node
• Improved service topology virtual node increase switch

Fixes¶

• Fixed display style of pod running status
• Fixed When the link feature is not enabled, hide related configuration parameters
• Fixed some front-end styles do not take effect
• Fixed collection pod metrics no data problem
• Fixed an issue of unable to install insight-agent in OpenShift cluster

2023.02.27¶

v0.14.6¶

New Features¶

• Added a refresh button to the chart
• Added service topology supports selecting multiple clusters and supports searching by service name
• Added service topology details and metrics of traffic egress and ingress
• Added a behavior of clicking the service name in the service topology details to jump to the details of the service
• Added list of service monitoring and metrics of traffic egress and ingress
• Added system component monitoring list
• Added CoreDNS monitoring panel
• Added an option to enable kubeAudit collection settings when installing Insight Agent

Improvements¶

• Improved filter conditions in trace query and view traces with Error
• Improved The scatter diagram of the trace query is updated to a bubble diagram
• Improved reduce Prometheus's metrics retention time to 2 hours
• Improved The default parameter of retentionPeriod of VMStorage is adjusted to 1 month
• Upgraded fluentbit helm chart version to 0.24.0
• Updated tailing-sidecar/operator mirror
• Updated Global collection rule interval is 60 seconds

Fixes¶

• Fixed built-in vmcluster dashboard
• Fixed When the link is not opened, the navigation bar cannot be loaded
• Fixed system component jump to view details link error
• Fixed The quick installation/uninstallation link of the acquisition management list is wrong
• Fixed After advanced index query query, the overlapping part of the index association and chart in the drop-down box cannot be selected
• Fixed Allow input of decimals when modifying the storage duration of historical alerts
• Fixed Send multiple notifications when an alert rule generates multiple alerts
• Fixed configmap-reload mirror error for vmalert and vmalertmanager
• Fixed fluentbit for Insight Agent on ARM architecture

2023.01.10¶

v0.13.2¶

Fixes¶

• Fixed the problem that kubernetes-event-exporter image address in insight-agent is wrong
• Fixed filter alerts API by resource name

2023.12.30¶

v0.13.1¶

Fixes¶

• Fixed build offline package to add .relok8s-images file
• Fixed adjust the port name corresponding to the component otel-collector port in insight-agent

2022.12.29¶

v0.13.0¶

New Features¶

• Added support for modifying the storage time of historical alerts
• Added status details of collection management components
• Added built-in message template
• Added chart metric Calculation Instructions

Improvements¶

• Improved log list field display
• Improved the judgment logic of insight-agent
• Upgraded Jaeger Chart version upgraded from v0.62.1 to 0.65.1

Fixes¶

• Fixed Some built-in alert rules do not take effect
• Fixed an error that the name can be duplicated when creating a rule
• Fixed DingTalk robot ending with '-'
• Fixed case-insensitive fuzzy search in alert rules
• Fixed service metric error delay calculation is not accurate
• Fixed Jaeger query has too many open files problem
• Fixed es index rollover alias and cleanup strategy not working

2022.11.28¶

v0.12¶

Features¶

• Added insight-agent Helm chart installation supports form

Improvements¶

• Improved PromQL queries support raw metrics
• Improved the style of the topology map
• Upgraded built-in MySQL image version, upgraded from v5.7.34 to v8.0.29.
• Upgraded Fluentbit ARM architecture helm Chart version from
• Upgraded kube-prometheus-stack helm Chart version upgraded from v39.6.0 to v41.9.1
• Updated used Bitnami mirror, including grafana-operator, grafana, kubernetes-event-exporter
• Updated prometheus-related API proxy address, change /prometheus to /apis/insight.io/prometheus

Fixes¶

• Fixed service list caching logic
• Fixed the problem that the built-in rules do not take effect
• Fixed request delay unit issue
• Fixed Insight internal link issue
• Disabled PSP resource in vm-stack
• Fixed victoriaMetrics operator not available in Kubernetes 1.25.
• Fixed browser compatibility issues with front-end mirroring

2022-11-21¶

v0.11¶

Improvements¶

• Added link troubleshooting and monitoring dashboard for component Jaeger
• Improved The alert list and message template list support sorting
• Improved to filter out clusters without insight-agent installed
• Improved sort by span start time by default when trace query

Fixes¶

• Fixed no data dashboard , including OpenTelemetry related dashboards
• Fixed the problem that there is no content under some log paths
• Fixed delete wrong alert rule: KubeletPodStartUpLatencyHigh

Other¶

• victoria-metrics-k8s-stack helm chart updated to v0.12.6
• opentelemetry-collector helm chart upgraded from v0.23.0 to v0.37.2
• jaeger helm chart upgraded from v0.57.0 to v0.62.1
• fluentbit helm chart upgraded from v0.20.9 to v1.9.9
• kubernetes-event-exporter helm chart upgraded from v1.4.21 to v2.0.0

2022-10-20¶

v0.10¶

Features¶

• Support for container-managed Service names associated with OTel service names to discern if service linking is enabled
• Updated the default tracking sample strategy in the global OTel column
• Change sumo (for audit logs) exporter port 8080 to 80
• Use go-migrate to manage database migration versions
• Fix multicluster and multi-namespace filters not working properly in graph API
• Support for building ARM images

Install¶

• Fluentbit supports docker and containerd log parsers
• Fix var/log/UTC issues
• Fluentbit supports elasticsearch output to skip TLS verification
• K8s audit log filter supports getting rules from Helm values
• Fix the parsing problem of centos7/ubuntu20 host log time
• Upgrade the OTel Operator version and remove the cert-manager dependency deployed with the self-signed certificate in the Operator
• Designed jaeger collector metrics
• Upgrade tailing-sidecar version
• Jaeger supports elasticsearch output to skip TLS verification
• Disable jaeger components in A-mode

Other¶

• Added OTel collector grafana dashboard
• Add Insight overview Chinese page

2022-9-25¶

v0.9¶

Features¶

• Support kpanda service name associated with the otel service name, identify whether the service tracing enabled.
• Update default tracing sample policies in global otel col.
• Change sumologic(work for audit log) exporter port 8080 to 80.
• Use go-migrate to manage db migration version.
• Fix multi cluster and multi namespaces filter not work well in graph API.
• Support build arm image.

Install¶

• Fluentbit support parser both docker and containerd log.
• Fix /var/log/ UTC issue.
• Fluentbit support elasticsearch output skip verify TLS.
• kube audit log filter support getting rule from helm values.
• Fix parse centos7/ubuntu20 host log time.
• Bump up otel operator version, remove cert-manager dependencies in operator deploy within self-signed cert.
• Scrape jaeger collector metrics.
• Bump up tailing-sidecar version.
• Jaeger support elasticsearch output skip verfify TLS.
• Disable jaeger components in Mode A.

Other¶

• Add otel collector grafana dashboard.
• Add Insight Overview Chinese version.

2022-8-21¶

v0.8¶

Features¶

• Migrate graph server into insight server.
• Add cluster_name param to graph query request.
• Add userinfo api.
• Add GetREDMetrics API in GraphQueryService.
• Add GetHelmInstallConfig api to get global cluster service addresses for agent to use.
• Complete auth module.
• Add init cmd/initcontainer for elasticsearch alias and ilm policy

Adjustment¶

• Bump up otel operator in agent chart.
• Add kibana as builtin tools.
• Reduce traces/logs chart's default values.
• Add Helm values parameters documentation.
• Polished Helm parameters.

Install¶

• Add audit log enable/disable feature.
• Move Fluentbit config to a ConfigMap.

2022-7-20¶

v0.7¶

Break changes¶

• Modify QueryOperations and GetServiceApdex's API definition in Tracing service.
• Remove resolve alert api.
• fix NFD master crash when CRDs missed.

Features¶

• Remove jaeger relate code in span-metric.
• Add index policy for skoala gateway logs.
• Add lua filter for Ghippo audit logs.
• Add global config api.
• Disable cache in vmselect component.
• Dock with ghippo roles.
• Expose metric insight_cluster_info in server.
• Add log.SearchLog API for SKoala, accept ES query DSL and return raw ES response.
• Bump up OTelcol helm chart version to 0.21.1 and update otelcol architecture.
• support mspider tracing.
• Bump up OTelcol helm chart version to 0.23.0.
• Add default tracing sample policies in global otel col.

Adjustment¶

• Use GrafanaOperator Stack to replace original Grafana Stack.
• Replace insight-overview dashboard.
• Add GrafanaDashboard, GrafanaDatasource CRDs.

2022-6-23¶

v0.6¶

Break changes¶

• Modify insight deployment and service name to insight-server.
• Modify trace relate metric query API response type.
• Using the unified paging mechanism

Features¶

• Add graph api through prometheus metrics of mesh layer.
• Add service graph api through prometheus metrics of general layer.
• Modify proto param, follow google style doc[https://developers.google.com/protocol-buffers/docs/style].
• Modify list api pagination and add sort.
• Add GProductVersion cr.
• Add insight metric config api.
• Manager insight license resource cr.
• Add traces api through access jaeger grpc endpoint with otlp/v2 protocol.
• Add service-detail api to get all metrics and scalars for a given service name.
• Add operation-detail api to get all metrics and scalars group by operation for a given service name.
• Add traces api through access jaeger grpc endpoint with jaeger v1 protocol.
• Add span metric protobuf style check

Adjustment¶

• Add node-feature-discovery subchart for License module.
• Add opentelemetry-collector subcharts to insight chart.
• Delete audit log OUTPUT config in fluent-bit.
• Add groupbytrace processor to generate trace/span number metrics.
• Add built-in Elasticsearch chart and enabled by default.

Install¶

• Upgrade victoria-metrics-k8s-stack chart version from 0.6.5 to 0.9.3.
• Add servicemonitor for components in victoria-metrics-k8s-stack.
• Modify insight components resource.

2022-5-18¶

v0.5¶

Features¶

• Add notification template API
• Complete rules and alerts API
• Add service API
• Add, delete, modify and check for vmrules
• Removed get query log API
• Support collecting kube audit logs with fluentbit
• Provide the function-related API of Service Graph
• Enhanced span_metric API: latencies, calls, errors three APIs support instance dimension query
• Enhanced span_metric API: Query Latency(with GroupByOperation) can return P99 P95 P90
• Enhanced span_metric API: latencies, calls, errors three APIs support extension_filters dimension query
• Added aggregated API for query latency, calls and errors
• Added apdex API
• Rename span_metric API URL

Install¶

• Added built-in mysql
• Upgrade GO version to 1.17
• Changed insight server service port from 8000 to 80
• Changed insight server/metrics port from 2022 to 81

Documentation¶

• Added documentation station glossary
• Added 4 pages of basic concept tasks and examples, data model, query language, etc. of the document station
• Added user guides - documents such as scene monitoring, data query, alert center, etc.
• New additions to the document site: Product Benefits, Metric Query, trace query, dashboard, overview

2022-4-22¶

v0.4¶

Features¶

• Increase the main API of the alert notification module
• Upgrade and adapt kpanda 0.4.x API
• Add the path information of the file to which the log belongs to the system log
• Add query single log context API
• Add query Kubernetes Event API
• Enhance Insight's own observability capabilities, provide its own metric interface and query trace information
• The API of Jaeger Query is used by the front end through the reverse proxy
• Add Query Tracing Operations related API
• Add Span Metric related API

test¶

• Added E2E use case coverage badge
• Supplementary test case documents related to alert notification
• Increase the E2E test of the log-related interface

Documentation¶

• Add overall bilingual document station structure and main content
• Added plug-ins required for documents to optimize rendering
• Completion of ROADMAP content
• Merge document ROADMAP content into total ROADMAP file
• Update document structure

2022-3-18¶

v0.3¶

Features¶

• gRPC and http use the same port
• Modify api path from /api/insight/ to /api/insight.io/
• Add cluster resource api proxy from kpanda
• ginkgo upgrade from 1.x to 2.x
• Organize proto files under /api
• split insight service in insight.proto
• Update kpanda api to 0.3.41
• Complete cluster/namespace list and cluster summary
• Add bulk query immediate and range metrics api
• add node and all workload api
• Added Otel tracing to track insight
• Support for metrics queries with extraLabels
• Add metrics documentation.
• Realize basic scenario cases in monitor

Helm Charts¶

• Add Jaeger helm chart
• Add OpenTelemetry collector helm chart
• Add tailing-sidecar-operator as an accessory/solution/plugin for log collection
• Add /variables/log/message collection in fluentbit
• Add kube exporter to collecot kube cluster event log

Comments