Changelog (2025)¶

December 24, 2025¶

Feature Updates¶

Role Permissions¶

Dashboard permissions refined: Added independent "Dashboard Create" permission. The original "Dashboard Management" permission still includes all operations. Assigning the "Create" permission will automatically enable basic dashboard and view viewing permissions.
Navigation bar dynamically displays based on user permissions. If a role has no permissions in a specific module, that module will not appear in its left navigation bar. This rule applies to the following modules: Cloud Billing, Billing, Scenarios, Event Center, Infrastructure, Logs, Metrics, APM, RUM, LLM Monitoring, Synthetic Tests, Monitoring, Incident, Security Monitoring.

Monitoring¶

Alert Policies:
- Simultaneously supports search and filtering for alert policy filtering; Newly supports precise filtering by monitor name (checker_title), member (member);
- Repeat Alert: Newly supports setting separate non-send intervals for repeat alerts based on different alert levels (e.g., critical, error, warning, etc.);
- New notification aggregation rule: "Custom". Supports freely specifying field combinations for alert aggregation based on business needs. Input field names (multiple fields separated by English commas), the system will strictly aggregate based on this combination: only when event data contains all specified fields and their values are valid will an aggregated alert be triggered.
Arbiter's built-in DQL functions support querying data from a specified workspace by passing a workspace parameter. Applicable to the creation and modification of Programmable Monitors and SIEM detection rules.

APM¶

Performance Metrics:
- Added Apdex statistics to quantify user satisfaction with application performance;
- Statistical metric time series visualization: Upgraded core statistical metrics in the service list (average requests, response time, error count, etc.) from numerical display to time series charts, intuitively showing trends.
Service Details:
- Supports displaying OpenTelemetry service languages;
- Added Tab pages: Service Map, Profile, Dashboard.

Scenarios¶

New chart type: "Change Comparison Chart": Quantitatively compares a specified metric's value in the current period with the same period in history (e.g., same time yesterday, same day last week) or other metrics, and intuitively displays the change magnitude for quick anomaly detection;
When creating a Pie Chart (Doughnut Chart), when the legend is not hidden, you can choose whether to display the Value and Value percentage.
Chart Y-axis: Optimized raw value display.

Snapshots¶

Overall optimization of the snapshot list style;
Save Snapshot: Added "Save as static time" setting.

Metrics¶

Metric Analysis: Numerical values in query results now support flexible display format settings, allowing configuration of decimal places and full precision result display.

Management¶

Cross-workspace Authorization: When the data scope selects "Metrics", the metric authorization scope can be refined;
Share Management: Added quick filtering by type and a search box.

Pipelines¶

Each data type now supports setting two default Pipelines (including local and central Pipeline).

Open API¶

Dashboard cross-workspace queries support specifying a default selected workspace.

Deployment Plan¶

Management Console:
- Workspace List > Tenant Configuration Management:
  - Added "Max Segment Time" setting, defining the maximum time search range for a single query. Once sufficient data is retrieved within this range, the search stops immediately;
  - Extended "Sampling Threshold" configuration, allowing free definition.
- Supports batch importing members by uploading a member list file;
- Security Settings > Login Method Management: Supports modifying the display name of the system default LDAP account to adapt to different users' internal unified login scenarios.
Added parameter (max_delay), supporting custom service topology calculation time.

New Integrations¶

Added GCP Storage integration;
Added GCP Cloud SQL (MySQL) integration;
Added Alibaba Cloud VPC Peering Connection integration;
Added AWS Lambda Trace integration;
Added Alibaba Cloud WAF integration;
Added Tencent CVM / KeewiDB views;
Fixed case sensitivity issues in the Ingress-nginx-prom integration documentation;
Fixed integration documentation issues (including Chinese/English character and format issues);
Fixed TencentCloud public CLB collector to support non-public IPs;
Fixed bug with linearBuckets metrics in GCP Cloud Monitoring;
Added CSPM rule: S3 bucket policies should restrict access from other AWS accounts;
Added cloud integration templates: Aliyun CPFS, Aliyun NLB, HuaweiCloud EVS, Aliyun CNAPIGateway, HuaweiCloud ROMA;
Added cloud script: GCP APIs collector;
Added cloud script: GCP RUN collector.

Bug Fixes¶

Fixed issue where alert events could not recover;
Fixed issue with excessive connections when connecting to external Kafka clusters;
Fixed issue where email addresses being blacklisted prevented normal email receipt;
Fixed issue where the initial load of the host list did not show all online hosts, requiring waiting for lazy loading;
Fixed issue where calling the log query task interface returned an error;
Fixed issue where clicking Mute in event details on the Hangzhou site did not take effect;
Fixed issue where the snapshot share button was sometimes unresponsive;
Fixed poor sampling results for APM on overseas sites;
Fixed issue where alert events could not associate and view related content;
Fixed issue where searching log sources triggered large queries, causing platform queries to return no results;
Fixed issue where high consumption warning emails were not received;
Fixed issue where monitor events did not recover as expected;
Fixed abnormal data display in the RUM Error Analysis view;
Fixed errors with trace data (7 days) and log queries (7 days);
Fixed issue with abnormal Sourcemap reporting for the same version;
Fixed issue where using wildcards for all indices when generating metrics from logs caused failures;
Fixed issue where read-only users could not view public dashboards;
Fixed issues with dashboard quick search not splicing with log stream graphs, and abnormal multi-select view variable display in log analysis;
Fixed issue where encountering errors when getting samples during timeline governance using pipelines on overseas sites;
Fixed incorrect time range (217 days) returned for the "Past 1 month" query range in overseas site scheduled reports;
Fixed dashboard template errors causing test failures on overseas sites;
Fixed abnormal expression calculation results in monitors;
Fixed frequent white screens and lack of event details after clicking the "Go to Workspace" link;
Fixed incorrect path in the guide for installing data collectors;
Fixed issue where user access geographic location displayed as Unknown;
Fixed abnormal CPU load value display for Windows hosts in Infrastructure;
Fixed abnormal results when using DQL for trace queries;
Fixed issue where Session Replay data could not be displayed;
Fixed incorrect APM data results after adjusting the time window on overseas sites;
Fixed issue where Inner Mongolia was displayed in English and data was incorrect in the China map;
Fixed issue where no results were returned after binding logs to external indices;
Fixed issue where trace data details could not be opened;
Fixed issue where pipeline processing logs could not judge/filter text with empty content;
Fixed IP address parsing city failures;
Fixed Chinese characters appearing in pipeline script functions on overseas sites, and issues with user IDs not displaying in the resource explorer, and dashboard errors in the resource analysis part of the user analysis board;
Fixed lag when variables select wildcards.

December 10, 2025¶

Feature Updates¶

Billing¶

Based on the newly added notification targets, added a new billing item "Phone Call": Counts daily voice notification sends (each/time per day).

In China region sites, billing details are as follows:

Billing Item	Settlement Currency	Price Details
Phone Call	RMB	￥0.6
	USD	$0.09

Monitoring¶

Added support for Voice Call (IVR) as a notification target;
Monitors support cross-workspace data detection.

Management¶

Data Forwarding: After performing the delete rule action, the system sets the rule status to "Deleting". After processing is complete, the rule is automatically removed.

December 3, 2025¶

Feature Updates¶

APM¶

Service Map:
- Clicking a node now slides out the corresponding service details on the right, showing basic information like service name, team, contact, plus new charts for service request response time, error request distribution, and error log count analysis;
- Added action buttons: zoom out, zoom in, display full topology map;
- Added topology map configuration:
  - Users can manually configure whether to calculate node size based on "Request Count";
  - Users can choose the hover highlight range.
Performance Metrics:
- Added filter field configuration. Besides official default filter fields, users can customize up to 5 fields as filter fields;
- Added support for exporting the current performance metrics list to a CSV file.
Service Details > Resource Invocation:
- Added sampling hint: Displays a "Sampling Rate icon" when resource list and distribution chart queries trigger sampling;
- Upstream/Downstream Topology: Added "View Upstream/Downstream Calls" buttons, supporting analysis of the full call path starting from the current resource based on the current topology.
Traces > Service Call Relationships:
- Interaction optimized: Only draws call paths between different services, making dependencies clearer;
- Node information enhanced: Displays full-path service call relationships using a left-right structure, directly showing execution time, average response time, call count, and error information on each service node; Clicking shows the service's parent span to trace upstream call sources; If the service has errors, error details are displayed directly.

RUM¶

To adapt to the HarmonyOS ecosystem development, added support for Application Type HarmonyOS. Users can choose the HarmonyOS type when creating an application and get corresponding SDK integration guidance and configuration support for complete HarmonyOS application observability.
View Explorer page supports Obsy AI error analysis.

Event Center¶

Added Event Level: Fatal (df_status:fatal), defined as the highest level.

DQL¶

Added function histogram_auto(): Automatic histogram range aggregation, no need to manually provide bucket basis, automatically selects appropriate bucket range and returns statistical information about data distribution.
eval expression null value handling logic optimized, added OR operator null value determination rule: When the left query result is empty, automatically uses the right result for calculation, flexibly adapting to null value completion scenarios. Example analysis:
- (a or 0) / b: When query a result is empty, it is automatically treated as 0, finally calculating 0 / b;
- (a or b) / b: When query a result is empty, automatically uses query b's result, finally calculating b / b.

MCP Server¶

Added support for querying metrics, traces, RUM data in a standardized way.

Management¶

Data Forwarding: When deleting a rule, the rule name must be entered to confirm the operation;
Member Management: Supports assigning roles to teams. Team roles are automatically applied to team members as additional roles; members' final permissions include personal roles and team roles.

Scenarios¶

Dashboard:
- Added a "Filter" button for quickly adding global filter conditions;
- Original functions like "History Version", "Large Screen Mode", "Report Record" moved to the "Settings" dropdown menu.
Charts:
- Advanced Configuration: Added "Repeat Chart" option. When enabled, the system generates multiple charts with the same structure based on the various values of the selected view variable;
- Clicking a data point now supports Apply to Filter, Copy Group Label;
- Table Chart: Manually entered return count上限 increased to 2000.

Monitoring¶

Alert aggregation notification template optimized: In rule aggregation and intelligent aggregation notifications, displays the number of different events triggered within the current aggregation周期, helping users quickly identify the scale and impact scope of alerts.

New Integrations¶

Openresty added OpenTelemetry trace access method;
Added Zabbix integration, based on api\stream;
Added Alibaba Cloud Native API Gateway integration, Alibaba Cloud Function Compute integration;
Added Huawei Cloud EVS cloud disk integration;
Added Tencent Cloud WAF monitor;
Optimized Alibaba Cloud NLB view & integration documentation;
Optimized Alibaba Cloud SLB view;
Fixed iis integration view monitor;
Optimized Huawei Cloud ECS, CSS, Functiongraph views and monitors;
Optimized Tencent Cloud clb monitor/view;
Added Huawei Cloud DC/ Alibaba Cloud SLB monitors;
Removed Huawei Cloud GuassDB related integrations (Huawei Cloud removed related products);
Fixed Huawei Cloud ROMA script;
Added CSPM templates:
- S3 buckets should enable "MFA Delete";
- S3 bucket ACLs should restrict public access;
- S3 bucket content access should be limited to authorized principals;
- IAM users should not have the "AdministratorAccess" policy attached;
- IAM policies should be attached and managed at the group level.
Added Resource Catalog templates:
- Huawei Cloud: dds, functiongraph, kafka, obs, rds(mysql), rds_mariadb, rds_sqlserver
- Volcengine: postgresql
- Alibaba Cloud: nat
- Tencent Cloud: tdsql_c_mysql, mongodb, eip, cos
- AWS: iam_user
Cloud Integration templates:
- Added Tencent Cloud ckafka, es, redis, waf, etc.;
- Optimized Huawei Cloud: rds, postgresl, apig;
- Optimized AWS: WAF.

Bug Fixes¶

Fixed issue where email was registered but couldn't log into the corresponding workspace;
Fixed issue where log data couldn't be written to newly created spaces;
Fixed issue where unrecovered event list showed no data after pagination;
Fixed issue where monitor indicators showed self-built dial test nodes sent data successfully, but no data in the explorer;
Fixed issue where file paths were not correctly displayed in error details after successful SourceMap upload and parsing;
Fixed issue where global configuration saving was无效 in the performance profiling explorer;
Fixed abnormal auto-refresh function for single dashboard view mode in scenarios;
Fixed issue where space admin accounts couldn't create new spaces;
Fixed time interval inconsistency between clicking view button in dashboard and dashboard configuration;
Fixed abnormal billing timeline charging function;
Fixed occasional inability to return correct results with sidebar filter conditions in log explorer;
Fixed issue where scheduled reports couldn't be generated;
Fixed incorrect copy on the dial test studio page;
Fixed "No metric data query permission" error for external API reports;
Fixed "Internal Server Error" when accounts with data access restrictions tried accessing the log explorer;
Fixed interface error when viewing network topology in Infrastructure Network menu;
Fixed issue where iOS symbol table uploads didn't successfully parse stack information;
Fixed frequent false alerts for Infrastructure存活 detection V2 version (corresponding machines online without异常);
Fixed inability to search for specific logs in Chinese within the workspace log explorer;
Fixed intermittent missing or incomplete trace query results requiring page refresh;
Fixed incorrect upper limit value (should be 10) for consecutive trigger condition settings in monitors;
Fixed abnormal 90% reduction in log data forwarding volume since a specific date;
Fixed inability to select error_stack field in monitor group settings, while DQL could use it normally;
Fixed incorrect data returned by DQL queries in monitors, causing event alerts;
Fixed Infrastructure function failure on Ningxia site;
Fixed issue where drilling down into CPU metrics showed no data in relevant charts;
Fixed issue where members could display logs via traceid but searching keywords in logs returned no results;
Fixed issue where log monitors could save without data source configuration, but monitoring would be无效;
Fixed error when opening a view in a mobile application within a scenario;
Fixed abnormal increase in "Cross-region private network traffic fee" billing item on a specific date within the space;
Fixed issue where high consumption warnings set within the Hangzhou site space didn't send预警 emails normally;
Fixed issue where AWS related resource catalogs displayed in Chinese;
Fixed issue with data auto-scaling function when setting dashboard data units;
Fixed dashboard resource topology data display errors and performance profiling page loading issues on Hangzhou site;
Fixed abnormal log queries on Oregon site;
Fixed service nodes in resource topology having no upstream/downstream, and request count/p99 metrics being 0;
Fixed issue where dashboard scheduled reports didn't display data normally;
Fixed abnormal jump to traces in APM analysis board;
Fixed abnormal error count increase (nearly ten thousand per minute) for specific service logs in deployment version.

November 20, 2025¶

Feature Updates¶

Query Tools¶

PromQL > Range query mode now supports setting step size (step).

Scenarios¶

Text Chart supports introducing view variables and time variables, automatically parsing user-manually entered view variables #{fieldname} and time variables.
Funnel Chart: The number of nodes that can be added has been increased to 15.
In dashboards, explorers, query tools, a sampling rate icon will be displayed in the upper right corner when chart sampling is enabled.
- In query tools, you can click this icon to temporarily disable sampling.

Management¶

In blacklist configuration, "Filter" now supports on-demand filling.
SSO Management:
- The "Edit" button must be clicked before enabling/disabling operations can be performed;
- The "Edit Identity Provider > Role Authorization" page supports selecting custom roles.

November 12, 2025¶

Feature Updates¶

Billing¶

The billing unit price for "APM/LLM Trace" has been adjusted as follows:

Monitoring¶

Added Programmable Detection monitors, allowing detection rules to be written via scripts, covering detection needs for various system data types.

Obsy AI Agent¶

Added support for anomaly analysis. Based on input text, the Obsy AI Agent can intelligently integrate complex and chaotic information to generate clear, concise root cause analysis results for you.

MCP Server¶

Added support for accessing MCP Server (Model Context Protocol). As an open protocol, it provides a unified multi-model callable interface service for the Guance Obsy AI Agent, supports multi-client access, and accesses Guance's monitor, log, dashboard, and DQL query capabilities via API Key authentication.

Management¶

Role Management: Scenario configuration management / monitor permissions split.
- Scenario configuration management: Permissions subdivided into
  - Dashboard, view viewing
  - Dashboard management
  - View management
  - Note, explorer management
- Monitors: Split into monitor viewing and management permissions.
Field Management: Added "Attribute Classification" configuration item, including: system fields, business fields, others.
Workspace Settings:
- Added "Data Reporting IP Whitelist" configuration. When enabled, only egress IPs within the whitelist can report data normally; data from other sources will be intercepted.
- The original "IP Whitelist" configuration is renamed "Login IP Whitelist". After enabling the login (host) IP whitelist, only requests from sources in the whitelist can log in normally; requests from other sources will be denied access.

LLM Monitoring¶

Added support for collecting scores (model quality evaluation scores, descriptions, etc.) and displaying them.

Event Center¶

Change Events: After a new host is accessed, a new change event will be generated.

Scenarios¶

Time Series Chart - Analysis Mode added style switching and period-over-period comparison;
Charts新增 "Open in Metric Analysis" function. Seamlessly transition from any chart to the metric analysis interface, with the system automatically带入 the set query conditions.

Explorers¶

Log Explorer: Display fields also support filtering values;
All Explorers: Added field type icon display.

Log Index¶

Log Index: Added description and associated service configuration. When querying traces, the system will automatically associate matching service indices.
External Index: Added index description configuration;
Support switching between single/multi-select modes when applying indices in the explorer.

APM¶

Service Map: Added display of maximum service response time.

RUM¶

Explorer > Fetch/XHR tab page: The original "Request" column is renamed "Resource URL" and supports displaying the full URL, which can be toggled显示与否 via a button.

Cloud Account Management¶

Added search control, allowing fuzzy matching by keyword search.

New Integrations¶

Added Volcengine PostgreSQL integration;
Added Alibaba Cloud CPFS integration;
Added Alibaba Cloud ALB integration;
Optimized Alibaba Cloud NAT view;
Optimized Tencent Cloud Cloud Monitoring;
Optimized Alibaba Cloud Oceanbase dashboard;
Added Pipeline templates:
- AWS ELB
- Squid
- Kong
- CloudFront
- CloudTrail
Added cloud collector templates:
- Alibaba Cloud: cfw/cen/apigateway/mqtt/nas/OceanBase/ga;
- AWS: step function/ssm run command/waf2;
- Huawei Cloud: apic/apig/auto_scaling/cbr/cdm/css_es/dcaas, etc.;
Added CSPM templates.

Bug Fixes¶

Fixed issue where clicking log details in log stream graphs couldn't display context;
Fixed issue where index auto-matching didn't work in log slide-out pages;
Fixed issue where too many dashboard view variables couldn't be selected;
Fixed issue where the latest DK version couldn't be obtained via Helm in AWS Oregon region;
Fixed DQL parsing failure when configuring detection metrics in APM metric detection;
Fixed workspace invitation member failure reports and language display issues on the APM interface;
Fixed display异常 when clicking the "Data Access" menu;
Fixed issue where selecting one-time "Immediate" report in Singapore site scheduled reports didn't receive emails;
Fixed chart jump issues in RUM performance analysis board;
Fixed page errors when viewing basic machine information on Oregon site;
Fixed issue where log sources under selected indices couldn't display normally in log detection monitors;
Fixed unresponsiveness when clicking log sources in monitors;
Fixed excessive connections when forwarding data to Kafka;
Fixed Internal Server Error when querying last day's data in APM analysis board;
Fixed duplicate display of extended field values in the console;
Fixed display异常 after selecting "All Sources" for data source in blacklist configuration;
Fixed abnormal trace information display in log details;
Fixed charts showing service unavailable in scheduled reports;
Fixed display异常 in APM > Resource Call > Response Time Distribution;
Fixed inability to set aliases for bar charts using PromQL queries for a metric without aggregation or grouping;
Fixed errors after clicking host details in the deployment version infrastructure page;
Fixed single group deletion errors in dashboards;
Fixed abnormal return results after changing top values in leaderboard chart settings;
Fixed blank and undownloadable Profile data;
Fixed user alerts caused by slow Redis hgetall queries;
Fixed label_values function being unavailable and view variable settings not saving after upgrading base version to 1.119.212;
Fixed incorrect jump link opening in security monitoring rule writing guide;
Fixed errors when querying external index logs;
Fixed excessive起伏 in data points in log query time series graphs;
Fixed empty configurations when exporting combined detection monitors from space A and importing to space B;
Fixed display and drill-down异常 in Heat maps.

October 22, 2025¶

Feature Updates¶

Open API¶

Supports modifying data retention policy duration;
Supports global log field configuration > key field configuration;
Supports creating log indices via interface.

External API¶

DQL¶

Added function rate_over_sum(), which first sums the field within the aggregation window, then divides by the window length (seconds), used to calculate the average rate, result is float type.
Added function rate_over_count(), which counts the number of non-null values of the field within the aggregation window, and divides by the window length (seconds), used to calculate event frequency, result is float type.
Added support for count(*) / 1i syntax.
Example: L::*:( count(*) / 1i ) [::1m]. Represents dividing by the time window size (1m) in seconds, calculating the QPS of log writes.

Billing¶

Data Forwarding > External Storage Forwarding Traffic Statistics: Based on the current workspace, summarizes compressed forwarding traffic by external storage type, counts daily increments, and distinguishes between public and internal network traffic for billing. Click to view specific price details.
Usage Analysis: Supports manually switching to the billing timezone (UTC+8), solving the issue of user timezone and billing timezone inconsistency.

Monitoring¶

Alert Policies: Added support for not sending notifications for recovery events: Enabling this option means all recovery conditions only generate events and do not send external notifications.
Monitors:
Added Large Volume Alert Protection button: When the number of alerts generated by a single detection exceeds a preset threshold, the system automatically switches to a summary-by-status strategy:不再 processing alert objects individually, but generating a small number of summary alerts based on event status and pushing them.
Log Detection > Detection Metrics: Supports configuring multiple indices.
Range Detection, Outlier Detection: Supports cross-workspace query of detection metrics.

Scenarios¶

Histogram:
Added support for metric data source queries.
Controls statistical dimensions based on the BY field in the query condition, divided into "Global Statistics" and "Grouped Statistics":
- No BY condition: Performs global total statistics, calculating the total distribution of all data.
- With BY condition: Groups by the BY field and then calculates the distribution of each group's data.
Added X-axis, Y-axis configuration, allowing custom data display boundaries by setting maximum and minimum values.
Time Series Chart: At the target time point on the time series chart, clicking allows selecting a fixed time marker. After setting the marker, it won't change due to mouse movement, page scrolling, or switching time series charts.
Combined Chart: Added export button, supporting one-click export of all chart information.
Chart > Link: When creating a new chart in a dashboard (existing charts remain unchanged), if the BY field in the query condition meets specific criteria, the system will automatically enable the display of the corresponding link list.
Reports: Original real-time reports and archive reports unified into "Reports". All reports are archived and saved, simplifying management and viewing experience.
Synchronously supports directly creating reports on the dashboard page, automatically带入 the current view and time range.
The Report Record page allows viewing historical archive data and directly sharing or deleting.

Infrastructure¶

Hosts: Added Real-time Performance switch. When enabled, the system refreshes every 10 seconds and automatically fetches the latest three metric fields: CPU usage, memory usage, and CPU load.

Explorers¶

Global Configuration > Quick Filter Configuration: For RUM, Synthetic Tests explorers:
Displays source list on the left, allows modifying multiple options simultaneously in a popup.
When opened from different explorers, the corresponding source is automatically selected.
RUM > Long Task Explorer > Details Page: Added Script details, showing还原 code.
Time Widget: Added Play button. After selecting a time range and executing the query, clicking the ▶️ button quickly returns to the most recent specified time range.

Snapshots¶

Supports re-editing already saved snapshots.

Management¶

Blacklist: Supports one-click enable/disable.
Pipelines: Optimized template export format and updated official template library samples.
Role Management: Supports one-click export of roles.
Data Forwarding: Besides JSON text format storage types,新增 supports Parquet data storage type.

LLM Monitoring¶

Added access guide configuration steps.

New Integrations¶

Added AWS CloudTrail integration;
Added AWS CloudFront integration;
Added AWS Athena integration;
Added Alibaba Cloud NAS integration;
Added Alibaba Cloud MQTT integration;
Added Alibaba Cloud OceanBase integration;
Optimized Kong integration documentation, added Tracing integration method;
Optimized CSPM templates.

Bug Fixes¶

Fixed issue where members couldn't view after log field advanced configuration linkage, and timestamp variable settings were incorrect;
Fixed issue where calling OpenAPI to invite members still sent invitation emails to already joined members;
Fixed issue where data in dashboard views wasn't expanded normally;
Fixed incorrect format after exporting table charts to CSV;
Fixed defects in the Multi-step test creation flow;
Fixed discrepancy between the return upper limit restriction for area charts, line charts, bar charts in metric analysis and the documentation description;
Fixed issue where Profile couldn't open normally;
Fixed issue where field aliases were visible when adding quick filters in the explorer but disappeared after adding;
Fixed scheduled report sending failures;
Fixed issue where no dial test task data was on the platform after creating a new dial test task, and self-observation queries found relevant error logs in kodo;
Fixed errors when querying JSON format data with DQL;
Fixed no data in RUM Action, View details;
Fixed issue where actual role permissions existed but weren't displayed on the right in API Keys management;
Fixed issue where scheduled reports in production space couldn't be sent, but test space could;
Fixed inability to delete custom roles;
Fixed issues in log explorer: including abnormal bar chart and data count display, and inconsistency between the number of items found搜索 and the number of items查出 below;
Fixed no data when viewing shared snapshots;
Fixed APM data empty result issues;
Fixed data堆积 when forwarding to Kafka, backuplog service consuming memory until OOM;
Fixed issue where Chinese map in dashboard didn't show provinces with data;
Fixed text charts moving to the top after refreshing after being moved in the panel;
Fixed slow loading of 5-minute log queries;
Fixed incorrect escaping of filter conditions when jumping to associated views in event details page;
Fixed issue where time window parameters affected calculation results when using DQL to query data;
Fixed guancedb-storage storing data for 1 year, not meeting expectations;
Fixed ineffectiveness of filtering user-defined fields in Actions;
Fixed issue where modifying Webhook object address didn't take effect normally in alert events;
Fixed log query issues: including lag when pulling down the list for more details and query errors, and significant differences between status distribution chart display results and actual data;
Fixed issue where cross-workspace trace queries only showed current workspace data.

October 15, 2025¶

Feature Updates¶

Chart Query¶

Expression Query:

Added a separate "Expression Query" input box for directly writing expression formulas;
Automatically assigns a query identifier (e.g., a, b, c) before all queries;
Compatibility Notes
Historical chart subquery identifiers (e.g., A, B) cannot be reused by new expressions;
Original expressions can still be queried normally,不影响 existing chart functionality;
It is recommended to use new query identifiers to reconfigure expressions when editing charts.

Data Forwarding¶

Added support for forwarding data to Volcengine TOS, providing five storage type options: Standard Storage, Infrequent Access Storage, Archive Flashback Storage, Archive Storage, Cold Archive Storage, to适应 different business scenario storage needs.
When the storage type for forwarded data is "Archive", supports restore tasks to thaw the data for access and analysis.

September 24, 2025¶

Feature Updates¶

Open API¶

Added specified API Key creation permissions for newly created workspaces;
Added support for binding specified billing centers;
Added support for cross-workspace authorization.

Security Monitoring¶

Added CSPM detection type, supporting custom selection of detection types based on actual business scenarios (like cloud resource protection, compliance auditing, etc.), enriching detection coverage scenarios. Detection type descriptions如下:
CSPM (Cloud Security Posture Management): Identifies misconfigurations (like open high-risk ports, excessive permission assignments) and compliance risks in cloud infrastructure, continuously assesses cloud resource security, reducing vulnerability exploitation risk.
SIEM (Security Information and Event Management): Collects and analyzes security data from the entire infrastructure, detects threats and anomalous activities through event correlation and log analysis (e.g., abnormal login + bulk data export).
Official Detection Library Experience Upgrade: Detection rules新增 "Belongs to Detection Library" identifier,便于 users quickly identify rule source; simultaneously supports switching views in the template library, allowing browsing all detection rules including monitors at once, improving rule查找 efficiency.
Security Detection Rule Editing Experience Optimized: Added "Trigger Detection Immediately" function,无需 waiting for scheduled detection cycles, can trigger detection tasks in real-time,便于 users quickly verify script effectiveness after writing/adjusting detection scripts, and immediately view generated security signals.

Management¶

Added Preferences, allowing unified configuration of workspace language, theme mode, detail page opening method, etc.;
Member Management: Added support for exporting members;
Invite Members: Added configuration for member teams and remarks.

Monitoring¶

Monitors: Supports "Trigger Detection Immediately" when editing monitor rules;
Mutation Detection, Range Detection V2 support cross-workspace queries
Mute Rules support cloning.

APM¶

Service list page added "Add Service" button.

Synthetic Tests¶

When creating new dial test tasks, supports configuring task execution frequency by inputting custom crontab.

Incident¶

When issues are automatically discovered for the first time or manually created, the system supports AI intelligent analysis, automatically summarizing the current issue content.

Metrics¶

Generate Metrics: Supports infrastructure, resource catalog data sources.

View Variables¶

Removed the original all values option, unified through * to achieve multi-scenario filtering needs, supports selective use based on different business scenarios, specific effects如下:

Ignore this field filter: When * is selected, ignore this field filter, chart queries will not include this variable's filter condition.
Full value filter (equivalent to original all values): When * is selected, all variable values will be used as filter conditions (supports up to 100 values). Note: Too much data may affect performance,建议 selective use based on actual needs, avoid indiscriminate full filtering.

Explorers¶

Log Explorer: Index style adjusted, simultaneously supports one-click select all indices, pin indices, search and locate based on index name, etc.;
All Explorers: Support Advanced Linkage Configuration.
Added query acceleration configuration, can configure indices for commonly used filter fields in the explorer "Global Configuration" to improve query speed for that field.

Added support for Quick Zoom function: One-click expands the current query time range to 3 times the original, reducing manual time range adjustment steps, improving data exploration efficiency.

External Data Sources¶

Added support for SQL Server data source.

New Integrations¶

Added AWS AOSS integration;
Added AWS PCS integration;
Added Alibaba Cloud CEN integration;
Added Continuous Profiler JVM dashboard template (analyzing Profile data);
Added AWS WAF integration;
Added FastDFS integration;
Added Alibaba Cloud API Gateway integration;
Added Alibaba Cloud Cloud Firewall integration;
Translated AWS AOSS/AWS PCS views.

Bug Fixes¶

Fixed issue where central Pipeline added labels to df_label in event content;
Fixed unsuccessful workspace authorization deletion;
Fixed resource catalog database resource display issues;
Fixed issue where host data decreased on the Infrastructure > Hosts page after clicking sort button;
Fixed incorrect billing center username取值 when registering new accounts;
Fixed system internal errors when binding external indices (TLS/ES);
Fixed errors when creating workspaces;
Fixed page errors reporting failed Func query task creation when drilling down into Redis metric data;
Fixed intermittent Internal Server Error when querying data in Dashboard;
Fixed "View not found" issue when jumping from container_crash related events in Infrastructure K8s;
Fixed identical interface display for different version prometheus exporter metric collection in Integrations;
Fixed missing display of some fields in log stream graph unit settings;
Fixed Open API call results not matching documentation;
Fixed lack of charts in system view "APM P90 Resource Analysis View";
Fixed alert policies failing to successfully send emails to customers;
Fixed JSON key:value format output from event templates not following the format viewed in the explorer;
Fixed false alerts generated by host data reporting monitors;
Fixed 502 errors when uploading SourceMap;
Fixed omissions in the Webhook alert process;
Fixed abnormal fields appearing when using DQL to query logs during panel drawing;
Fixed unknown Scalar function error in dashboard queries;
Fixed mismatches after switching DQL to simple query in dashboards;
Fixed Pie Chart in Studio dashboard > Data Display module not working properly;
Fixed 400 error on the Hangzhou site, metric management interface > Timeseries page;
Fixed incorrect copy in Integration > Cloud Account Management;
Fixed errors in RUM ERROR pattern analysis;
Fixed inability to query Session and View data in RUM;
Fixed Warroom issues;
Fixed no results in the list when filtering span types for entry and exit types for a specified resource;
Fixed incomplete service list display.

September 3, 2025¶

Feature Updates¶

Security Inspection¶

The Security Inspection function is officially下线. Its related functions can be achieved through the Security Monitoring > Security Event Management (SIEM) module.

LLM Monitoring¶

Guance launched LLM Monitoring, providing "monitoring + tracing" capabilities for large model applications,告别 model operation "black box", completely recording input/output, tool calls, retries, latency, cost, and tokens for each call, using visual Trace and Session aggregated by session for easy playback and troubleshooting, achieving full-link transparent management. Now优先 supports Langfuse integration. If you are already a Langfuse user, you can quickly enable LLM monitoring capabilities.

Free Plan¶

面向 independent developers and small teams, Guance新增 Free Zone (Beijing) node. Under this node, you can register the Free version (other sites no longer support creating Free Plan workspaces), covering key core functions.

Open API¶

Incident > Schedule: Supports adding parameters for getting calendars and filtering schedules.

Explorers¶

Log Explorer:
Added Advanced Linkage Configuration: Supports configuring jump links for individual fields, allowing direct opening of target pages from the current field, completing parameter transfer through variables for linkage analysis.
Log Explorer Display Items optimized:
- Added "Delete All Fields", "Reset Fields" buttons;
- Added "Time Column" display switch.
All Explorers
Supports DQL query mode.
Supports searching all IP addresses under a single IP field in event content via the CIDR function.

Event Center¶

Events generated by SIEM rules (security) support Obsy AI intelligent analysis on the details page.
No restriction on dimension tags,默认 supports jumping to log/trace explorer.

Monitoring¶

Alert Policy Management: Under rule aggregation mode, supports selecting only "Send First Alert", thus avoiding missing important异常 events while waiting for aggregation.
Monitors > Event content supports defining domain variables:
df_studio_env_name: Current environment name
df_studio_console_base_url: Current environment console address
df_studio_monitor_base_link: Current environment monitor address

Management¶

Workspace admins support turning off AI intelligent analysis at the workspace level.

Metrics > Metric Analysis¶

Table Chart mode新增 Prom mode, displaying PromQL query return results.
In independent chart mode, chart titles are added based on query conditions.

Infrastructure¶

Host, container explorers support directly creating mute rules.

Synthetic Tests¶

Dial test tasks: Request body content新增 supports multipart/form-data.

Sensitive Data Scanning¶

Supports covering all data types, no longer limited to scanning only log data.

Mobile RUM¶

Analysis board新增 "Crash Analysis", "Stutter Analysis" tab pages.

New Integrations¶

Added Squid integration;
Added Openresty integration;
Added Kafka 4.0 integration;
Added AWS MWAA integration;
Added AWS Step Function State Machine integration;
Added Alibaba Cloud GA integration;
Updated Nginx tracing integration;
Updated AWS OpenSearch monitors/views.

Bug Fixes¶

Fixed issue where aliases in dashboard grouped table charts were ineffective;
Fixed incorrect overview chart display results;
Fixed Chinese characters appearing in Import Dashboard (multiple dashboards - zip file);
Fixed issue where funnel chart跳转 to session list, but session replay displayed errors;
Fixed alerts still being sent despite mute settings;
Fixed host tags not displaying in infrastructure;
Fixed charts in the analysis board not displaying data for the corresponding time period even though data existed in the Android端 explorer;
Fixed lack of connections between services in the service topology map;
Fixed incorrect alert mute configuration causing all alerts to be无效;
Fixed DQL statement queries returning no results in charts;
Fixed dial test task errors on the Guangzhou site;
Fixed unexpected results after importing templates in the deployment version management console.

August 13, 2025¶

Feature Updates¶

Open API¶

Supports custom roles for Tokens.

RUM¶

To cover query scenarios for application data of different volumes, the system optimized RUM data writing. This同步 leads to the following updates:

RUM data query drills down to the application level:
Explorer/Analysis Board: Added application filter at the top of the page,默认 selects "All Apps", supports custom single or multiple selection; query is triggered only after selecting an application.
DQL query statements support single or multiple application queries; more usage instructions can be found at RUM Data.

R(app name1)::view:(count_distinct(username))   # Specify querying UV for a specific application name
R(app id1)::view:(count_distinct(username))     # Specify querying UV for a specific application ID
R(*)::view:(count_distinct(username))   # Specify querying UV for all applications in the space

Existing users experiencing application data query performance bottlenecks can contact their account manager to enable the "User Access Data Split Table" function in the management console. For new users, after creating a workspace, the system will automatically split table data writes based on different applications.

Scenarios¶

Scheduled Reports: Added "Archive Reports": Supports full data calculation for dashboards according to set cycles (daily/weekly) and generating static snapshots. Archive results are retained long-term, unaffected by dashboard configuration changes, and can be viewed within the platform or notifications sent as configured.
Dashboard > Log Analysis: Added support for automatically带入 the dashboard's view variables as filter conditions for display.
Log Stream Graph: Supports querying all data sources within the platform.

Explorers¶

Search bar supports querystring(md5(xxx)) tokenized query method. If data is MD5 encrypted and needs to be located in the explorer after scanning, this method can be used for searching.

DataFlux Func¶

To enhance security protection for DataFlux Func Hosted Edition, the following updates were made:

Removed the "Reset Password" operation from the Hosted Edition page in the console, and stopped sending initial account passwords to email;
For newly created Hosted Editions, logging in via the Func page is disabled by default, only supporting免登 via console user authorization.
Installed Hosted Func: After upgrading to version 7.2.0 or higher, supports two login methods:
Authorized Studio users can log in directly through the Console > DataFlux Func Hosted Edition interface;
Users previously created in Func (including the initialized admin user) can also log in via the Func page.

APM¶

Traces > Waterfall Detail Page/Code Hotspot tab page: Resources support one-click expand or collapse;
Topology Map: Added support for "P95 Response Time" display.

Synthetic Tests¶

Dial Test Tasks > Request Body: Supports configuring four types of dynamic variables:

{{ timestamp "ns" }} //1754379375400790109
{{ date "rfc3339" }} // 2006-01-02T15:04:05Z07:00
{{ date "iso8601" }} // 2006-01-02T15:04:05Z
{{ urlencode "hello world" }} // hello+world

Added support for one-click cloning of dial test tasks.

Monitoring¶

Threshold Detection > Detection Metrics: Supports selecting metric data across workspaces. After the monitor rule is created, cross-workspace alert configuration can be achieved.
Monitor List: After quick filtering monitors, supports cross-page select all and one-click export.

Events¶

For recovery events, added definition of df_fault_end_time extended field as fault recovery time.

DCA¶

List supports selecting all data;
Supports configuring filter conditions based on DataKit version (datakit_ver), operating system (os), environment (env), etc.;
List新增 environment display.

Management¶

Role Permissions: Added "Client Token View" permission.
Cross-Workspace Authorization: When editing or deleting authorization rules, if data scope or log indices are reduced, a confirmation popup appears after confirmation, displaying the names of affected monitors and data access configurations, simultaneously supporting exporting this list as a CSV file.
System Notifications: Added subscription type – "Monitors": Once a monitor becomes无效, the system will send email notification.

New Integrations¶

Added AWS SSM RunCommand integration;
Added AWS EBS integration;
Added AWS ECR integration;
Added AWS EFS integration;
Added TDengine monitor;
Added Milvus K8s view;
Updated Azure cloud bill: EA architecture support;
Rewrote OTEL jvm dashboard: different metrics for 1._ and 2._ (java agent);
Huawei Cloud bill script update: Fixed billing异常 caused by missing instance IDs;
SIEM templates launched (10 AWS templates);
Optimized AWS MemoryDB;
Optimized Tomcat monitor;
Optimized Tencent Cloud MySQL, supports cloud disk edition (cloud disk) node dimension metrics, slow log collection;
Optimized Alibaba Cloud Oceanbase collector, supports supplementing tenant data.

Bug Fixes¶

Fixed recent slow log queries;
Fixed inconsistent data performance of resource objects;
Fixed trace associated logs显示 as 0;
Fixed errors when clicking code hotspots after successful deployment;
Fixed abnormal data display when sliding to the bottom of the dashboard log stream graph;
Fixed import errors caused by setting identifier when importing dashboard JSON;
Fixed incorrect jump locations in the resource catalog official template library;
Fixed sensitive data scanning configuration not taking effect;
Fixed issue where count_distinct query results were more than count query results;
Fixed no data display on the trace details page;
Fixed PromQL statement errors, which didn't occur when changing the query time interval;
Fixed trace service name not displaying normally;
Fixed unmodifiable Picture charts;
Fixed inability to write cloud bills in production environment;
Fixed abnormal service topology chart display in dashboards;
Failed site registration and activation;
Fixed abnormal RUM funnel function display;
Fixed intermittent missing dashboards after importing dashboards, requiring re-entering the page to display completely;
Fixed misleading interface access guidance;
Fixed interface operations reporting 406 - kodo API异常;
Fixed SourceMap file path, environment, version all correct but unable to parse source code;
Fixed abnormal log index data display;
Fixed APM guide page addition on the page;
Fixed variable display in chart titles after sharing charts;
Fixed service context not updating when clicking different services in the trace flame graph;
Fixed abnormal fuzzy query search data for dashboard view variables;
Fixed Pipeline script save failures;
Fixed SEIM templates not taking effect after addition, but labels生效;
Fixed topic 404 in deployment version nsqd logs;
Fixed inability to normally display loading time in popular page performance overview;
Fixed no view details button for .net profile data on the trace details page;
Fixed intermittent no data display for the same DQL expression query;
Fixed unstable infrastructure dependency display in APM;
Fixed DQL subquery ineffectiveness;
Fixed platform page errors after muting events in the event center.

July 23, 2025¶

Feature Updates¶

Monitoring¶

SLO
SLO deduction executed at "second" level frequency;
Added display of corresponding event information generated under the monitor in SLI events;
Details Page新增 overview page, relevant key metric charts dynamically统计联动 with the time widget in the upper right corner.
Added "SLO Correction" configuration. When SLO is about to be violated due to uncontrollable factors (e.g., infrastructure failure, third-party service interruption), temporary correction mechanism avoids false alarms while ensuring SLO data audit compliance.
Alert Policies
Supports cloning alert policies;
When configuring notification rules based on members,新增 supports one-click copying member notification configuration.

Event Center¶

The default explorer page's time widget refresh time is set to refresh every 30 seconds by default;
Alert notifications on the details page no longer display sending status.

Billing¶

Usage Statistics: Added "Backup Log Increment" display, used for counting daily新增 backup logs;
Usage Analysis: Default entry query time adjusted to "Today".

Basic Components¶

For the Shortcut > Snapshot display page:
Supports filtering by type;
Supports batch delete operations.
Explorer
Search Bar:
- Recent query display count adjusted to 10, save count adjusted to 50;
- Supports pinning query conditions;
Quick Filter > Space filter item: After configuring fields, the system automatically adds indices. Simultaneously, in the explorer search bar and chart query dropdown components, an icon identifier is added for these fields.
Details Page: Extended fields新增 icon identifier for fields already added to the display column, for easy distinction.

Infrastructure¶

Database: Added SQL Server data collection type. The details page can view all Table related information, including columns, indexes, foreign keys.
Resource Catalog: Custom object explorer details page新增 "Complete Information" tab page.

Metrics¶

Metric Management: Added label "Unique Value Count" display,统计 the distinct count of label values globally or under a single metric dimension.

Pipelines¶

One-click get sample: Supports specifying data sources to get test data.

Scenarios > Leaderboard¶

External data source queries support sorting.

Management¶

Data Forwarding: Supports OR for multiple filter conditions;
Token: When replacing Token,新增 configuration to adjust permission scope, supporting use only for reporting.
Sensitive Data Scanning: Custom rules support lookaround regex test matching;
Support Center > Ticket: Added ticket description character limit.

New Integrations¶

Added Huawei Cloud APIG integration;
Added Huawei Cloud SFS Turbo integration;
Optimized Alibaba Cloud ECS integration view;
Updated AWS SQS integration;
Tencent Cloud cloud product collection optimization, supports multi-dimensional metric collection (Cloud Monitoring V2);
EIP
COS
MySQL
PostgreSQL
MariaDB
TDSQL-C MySQL
TDSQL-C PostgreSQL
Redis
KeeWiDB
MongoDB
Memcached
Huawei Cloud integration related documentation content format adjusted.

Bug Fixes¶

Fixed issue where manually changing the status of an Issue created by auto-discovery prevented new Issue creation;
Fixed abnormal log explorer status distribution chart display;
Fixed expired session replay display;
Fixed no data after switching event center display items to list mode;
Abnormal event data frontend display;
Fixed abnormal RUM explorer Resource data display;
Fixed query errors for forwarded data;
Fixed /api/v1/inner/alert_opt/get, /api/v1/account/list interface exceptions in deployment version;
Fixed outdated Pod monitoring view template associated with traces;
Abnormal data access input data scope filter display;
Fixed time range issues when slicing PromQL by 1d in time series chart > bar chart;
Fixed PromQL query data existing in hourly intervals but not in daily intervals;
Fixed webpack plugin showing SourceMap file upload success but unable to view in SourceMap list;
Fixed duplicate Issues generated by Incident > Issue auto-discovery;
Fixed Prometheus function idelta not meeting expectations;
Large gap between log explorer query results and DQL query results under the same query conditions;
Fixed monitors setting df_label but events generated had no value for df_label, causing user mute policies not to生效;
Fixed IP field in Pipeline for user access not生效;
Fixed errors after kodo-x upgrade in private environment;
Fixed trace details显示 "top-level Span missing", but top-level Span actually exists;
Fixed inability to query application data in RUM > Application List;
Fixed workspace billing not showing consumption details;
Fixed snapshot saving issue where pressing enter while entering the name would save directly;
Fixed log monthly bill usage being 0 but still charged;
Fixed display judgment for src_ip and dst_ip;
Fixed occasional error display in log explorer bar chart for the same time interval;
Fixed search function usage issues in Incident;
Fixed Synthetic Tests tasks勾选 follow direction not生效 and being cleared upon re-editing;
Fixed inconsistency between log index query count over multiple days and TLS.

July 9, 2025¶

Feature Updates¶

Billing¶

Introduced a new billing model: Enterprise Edition. This model uses a combination of data traffic + total storage volume billing, more accurately reflecting actual resource consumption in extremely high log write volume scenarios, making user cost structures more transparent and predictable. This mode offers optimized cost efficiency for handling massive observability data, serving as a foundational capability for building an "enterprise-grade observability hub with full-domain control."
Within Enterprise Edition workspaces, data storage policies support custom configuration;
To upgrade from Free Plan to Enterprise Edition, please contact your account manager.
Commercial Plan:
Billing item names in Usage Statistics and Usage Analysis adjusted;
Commercial Plan Timeseries 180-day, 360-day prices adjusted.

Monitoring¶

Monitors > Threshold Detection: Detection metrics support querying external data sources;
Monitor List: Supports searching and locating monitors based on monitor ID.

Event Center¶

Added Display Items: Supports switching between standard and list mode display.
Change Events > Details Page: Basic Attributes adjusted as follows:
Change Content: Displays df_message content, recording detailed descriptions of resource changes;
Difference Comparison: Displays diff content, recording the difference text of resource changes.

Scenarios > Charts¶

SLO Chart: Locked time新增 "This Year" dynamic time option;
Heatmap:
Added "Metrics" data source type query;
Added Y-axis configuration, supporting Y-axis calculation based on data.

Infrastructure¶

Network > Overview
Supports custom analysis dimensions: Displays aggregated query data by selecting source and target query fields;
Supports cross-workspace network data queries.
Database:
Added Oracle database integration;
For PostgreSQL databases, all Table related information can be viewed, including columns, indexes, foreign keys.

Metrics¶

Generate Metrics: Supports different generation rules specifying the same metric name and tags;
Metric Management: Added timeseries multi-dimensional analysis at the measurement level.

Deployment Plan Updates¶

Workspace List > Index Configuration: Supports configuring data storage policies.

New Integrations¶

Added OpenLLMetry integration;
Added Undertow integration;
Added Cloudflare integration;
Added Huawei Cloud CDM integration;
Added Huawei Cloud DWS integration;
Added Huawei Cloud NAT integration;
Added AWS Transit Gateway integration;
Updated Volcengine ECS monitors, views;
Updated MinIO views, monitors;
Optimized integrations: Huawei Cloud ASM / CCE;
Optimized integration: Alibaba Cloud SAE;
Optimized integrations: JMX / MQTT.

Bug Fixes¶

Fixed 404 error when viewing session replay data accessing SaaS environment via免登录 Token;
Fixed Issue auto-discovery configuration content not being recognized in RUM;
Fixed lack of specific关联 information display in event details for monitor dial test alerts;
Fixed prompt for existing duplicate label names after modifying label names added in global tags;
Fixed abnormal DQL query return results;
Failed PromQL metric queries, but normal DQL queries;
Fixed混乱 Deployment > Pod information display in multi-cluster scenarios;
Fixed abnormal log status distribution chart display;
Fixed RUM filter time range inconsistency with time selector, causing inability to view application data;
Fixed Issue auto-discovery template variables not生效 in RUM Error explorer;
Fixed "Invalid address" error when adding Webhook notification targets;
Fixed data堆积 in base forwarding;
Fixed abnormal log status distribution chart display (duplicate line, likely error);
Frequent false alerts caused by DQL query data being 0;
Fixed inability to sort, slimit无效, and paginate in base query tool log queries;
Fixed timeseries queries returning timelines without any tags and不受 filtering constraints.

June 25, 2025¶

Breaking Changes¶

Extracted df_check_targets, df_check_condition, df_check_rules from the df_meta field of events to the same level. Subsequently, in webhook notifications, when df_source = monitor, information like detection time, detection metrics, detection count, and detection rules from the monitor configuration will be displayed.

Feature Updates¶

External API¶

Added interface index_list_info to query workspace index information list;
Added interface update_index_setting to modify workspace index configuration.

External Data Sources¶

Added integrations for PG, Oracle, Clickhouse external data source types to meet broader data scenario needs.

Logs¶

Added Native Direct Write Index type, allowing specifying the write index during data collection for precise classification and storage.

Monitoring¶

Range Detection V2 supports detecting APM, RUM data types;
Mutation Detection > Event Content: Added df_check_detail extended field, supporting outputting the mutation comparison original value to event details.

Events¶

Alert events triggered based on monitor configuration rules support Obsy AI Intelligent Analysis.

Infrastructure > Database¶

Added PostgreSQL database integration;
For MySQL databases, all Table related information can be viewed, including columns, indexes, foreign keys.

Explorers¶

Search bar added Query History display, click to directly reuse and flexibly modify.

Cloud Account Management¶

Added support for Tencent Cloud, Azure account types.

Scenarios¶

Adjusted Import Dashboard entry, integrating the original custom template import function, simultaneously supporting batch dashboard import.
Dashboard page新增 "Log Analysis" function,无需 page jump, directly viewing and analyzing related log data in a slide-out new page with one click.

Incident¶

The "Incident" menu item on the console left side新增 "Incident Management" entry,可以直接 jump to the management page.

New Integrations¶

Added Volcengine NAT integration;
Added Ceph integration;
Updated Clickhouse integration;
Huawei Cloud cloud bill script optimization;
Added Tencent Cloud bill;
Added Tencent Cloud WAF integration;
Added Tencent Cloud ES integration;
Added Tencent Cloud cos/cdb/mariadb/postgresql English views, monitors;
Added AWS Firehose;
Added AWS VPC NAT Gateway integration documentation;
AWS cloud bill script update: T+1 calculation caused billing errors for some components (e.g., EC2), adjusted to T-2;
Added Azure cloud bill.

Bug Fixes¶

Fixed abnormal data display after generating metrics;
Fixed inability to load log content after clicking "Load More" on the log details page;
Fixed invalid RUM开通 link and lack of quotation content;
Fixed inability to query data after clicking associated event jump in Issue;
Fixed 406 error when creating Infrastructure存活 detection v2 type monitors;
Fixed slow session replay page loading;
Fixed Open API creating mute rules without auto-generating rule names;
Fixed errors when inputting PromQL to query metrics;
Fixed abnormal page error rate display;
Frequent disappearance of host tags;
Fixed降序排列无效 in resource invocation leaderboard after adding search conditions;
Fixed no prompt to re-bind after unbinding MFA and logging out重新登录 if 7-day免 MFA login was勾选;
Fixed inability to render correctly in exported dashboard CSV files due to empty aliases;
Fixed monitors not generating alerts normally, self-observation logs reported query timeout errors;
Automatic deletion of imported fields in background management;
Lost user experience data.

June 11, 2025¶

Feature Updates¶

Billing¶

Added statistical items: Trace write traffic, RUM write traffic, SIEM scan volume; Usage Analysis同步 displays新增 statistical item data.

Scenarios¶

Dashboard > AI Intelligent Analysis: Added support for automatic error discovery and error analysis, further enhancing Obsy AI data processing capabilities.

Monitors¶

Detection frequency supports shortest custom input of 20s, system新增 supports selecting 30s;
APM Metric Detection, RUM Metric Detection support multi-select service, application respectively;

SIEM¶

Official Detection Library launched: The system provides out-of-the-box security information detection template library.

Infrastructure > Database¶

Supports cross-workspace data viewing;
Added fields: QPS, TPS, Average Query Time, Slow Query Log Status, Slow Query Count.

Metrics > Metric Analysis¶

Added Independent Chart mode, suitable for scenarios where multiple queries are added simultaneously and the query result units are inconsistent;
Added support for displaying grouped total statistics, simultaneously default slimit changed to 50.

Logs > Explorer¶

Stack Mode新增 line break switch;
Context logs support clicking the Time column for ascending/descending sorting.

Cloud Account Management¶

Cloud Account type新增 supports Huawei Cloud.

Management > Environment Variables¶

RUM Android/iOS SDK supports dynamic updates via Environment Variables.

Query Tools¶

Table charts and metric analysis simple queries both support selecting multiple columns;
DQL quick query defaults to simple mode, supports multi-column configuration queries.

APM¶

ddtrace新增 package level埋点 configuration. If you need to埋点 all methods under packages com.zy and javax.servlet, you can configure it as follows: -Ddd.trace.method.packages=com.zy,javax.servlet.

Deployment Plan Updates¶

Mapping Rules新增 "Append Mapping" switch. When enabled, accounts using single sign-on will be assigned workspaces and roles according to mapping rules, while original workspaces and roles remain unchanged.
For workspaces using the ScopeDB engine, supports configuring tenant-level queries.

New Integrations¶

1..1. Optimized Tencent Cloud bill script, supports obtaining bill information via sub-account method; 2. Added Tencent Cloud bill integration; 3. Added Tencent Cloud EIP integration documentation; 4. Added Tencent Cloud CDB monitor (English); 5. Added Huawei Cloud cloud bill integration; 6. AWS ELB integration content updated; 7. AWS LAMBDA integration content updated; 8. AWS S3 integration content updated; 9. Added AWS Aurora MySQL integration; 10. Added AWS Aurora PostgreSQL integration; 11. Added Hikaricp integration; 12. TiDB Cloud view optimization; 13. Added vpc integration; 14. Added AWS rds postgresql integration.

Bug Fixes¶

Fixed issue where userid was not displayed in the resource explorer of Singapore site user spaces;
Fixed inability to query custom reported userid through the action explorer;
Fixed inaccurate dashboard AI analysis results;
Fixed Open API获取时序趋势图 issues;
Fixed delayed display of data gap content on the monitor modification page;
Fixed skipping existing files when uploading SourceMap using webpack plugin;
Fixed errors in quick filters under network functionality;
Fixed malfunction of variable filters;
Fixed high-risk vulnerabilities identified in penetration testing;
Fixed dashboard errors in the resource analysis section of the Singapore site user analysis board;
Fixed undefined variable values in panel link jumps;
Fixed missing time selection box in the upper right corner of the log explorer;
Fixed discrepancy between preview chart data and actual data in monitors (APM);
Fixed internal display issues regarding DingTalk and WeChat in cross-space authorization emails;
Inconsistent SSO user list and user details in the management console;
Fixed display issues with external database data source queries;
Fixed display issues with trend charts in new threshold monitors;
Fixed search failures when viewing services in APM;
Fixed DQL query errors;
Fixed daily repeat mute rules not sending mute生效 notifications as expected daily;
Fixed inability to view the AI analysis button after entering traces from RUM session replay.

May 28, 2025¶

Open API Updates¶

Incident supports returning associated event count information when listing and retrieving.

Feature Updates¶

SIEM¶

Detection Rules: Added test, full-screen script editing functions;
Added Signals explorer;
Added Execution Logs explorer.

Monitoring¶

Alert Policies > Association: Added selection of SIEM detection rules,同步 supports direct creation.
Editing Monitor Rules > Detection Metrics新增 supports PromQL query preview.
Monitor Rules List:
Importing新增 supports "Delete and import monitors with the same name" on the operation page;
Supports one-click export of all monitor rule configurations in the current workspace.
Intelligent Inspection function entry下线.

Infrastructure¶

Containers: Added Persistent Volume Claims Explorer.
Added Database Explorer, providing a centralized explorer tool for real-time monitoring of MySQL database performance metrics, resource usage, and log records.
Resource Catalog: Resource classification新增 database, and displays the total number of all databases.

Incident¶

Channel > Issue Details Page: Added "Associated Events",可以直接 click to jump to the event list for viewing; Issue emails and other sent notification events同步新增 associated event display;
Notification Policies > Issue Escalation: Added event storm option. Can set the system to send "repeat" or "once" notification reminders after the number of events associated with an Issue累计 exceeds a specific number.

Events¶

Unrecovered Events: - Added "Mute" button. Clicking allows direct "Mute" configuration of the rule on the current page. - Added "Recover All" button. Clicking recovers all abnormal events in the current list, with option to associate with Issues.

APM¶

Flame Graph: Added "Locate to Current Span" function button.

RUM¶

WEB SDK新增 client timezone recording.
Session Replay view page新增 client timezone display.
All time widgets统一 added timezone display.

Scenarios¶

Dashboard > Obsy AI Intelligent Analysis:
Optimized page display效果;
Optimized problem output.
Charts
Time Series Chart: Added hint: Query has exceeded the maximum number of time series data points, only displaying 1,000 data points;
Bar chart mouse hover color adjusted;
Table chart order adjustment optimized: Compatible with PromQL queries.

Metrics¶

Metric Analysis: Added query history and unit configuration.

Management¶

Sensitive Data Masking: Supports configuring masking for multiple fields with one rule.
Data Access: Edit page新增 access site selection.

Explorers > Quick Filters¶

All Explorers:
In Edit Mode, supports modifying "Filter Fields", "Optional Fields";
Added hide field alias function;
When searching for fields, if the query result has no exact matching value,可以直接 create and add to "Filter Fields".
Log Explorer:
Adjusted display of two view modes, added "Global Configuration" in "Stack Mode";
Display Item Configuration adjusted.

Registration¶

Login registration page display optimized.

Deployment Plan Updates¶

Configuration Management > Issue Discovery: Rule check frequency新增 supports 1 minute.

New Integrations¶

Added TiDB Cloud integration documentation;
Added Traefik integration documentation;
Added AWS cognito integration;
Added AWS certificatemanager integration;
Added AWS RDS PostgreSQL integration documentation;
Added aliyun rabbitmq integration;
Updated haproxy views and monitors;
Added Huawei Cloud DIS English monitors, dashboards;
Updated AWS ELB application integration documentation, views, added AWS ELB application monitor;
Updated AWS lambda integration documentation, added Lambda fetching S3 log data operation steps.

Bug Fixes¶

Fixed issue where generated metric data was not reported normally.
Fixed tag import failures when calling Open API import monitor interface.
Fixed unstopped dial test tasks.
Fixed 406 error when clicking "View Context Logs" in log details.
Fixed cache issue where alertPolicyMemberInfos displayed existing members in code when monitors only configured notification targets but not members.
Fixed issue where saving was无效 after selecting to treat results as 0 for data gaps in monitors.
Fixed failures and errors during batch alert recovery.
Fixed inconsistent log explorer query result counts.
Fixed overwriting original data when batch modifying monitors via API without configuring extend, rule fields.
Fixed incorrect replacement of built-in view templates after import in background management.
Fixed lack of linkage between log index configuration one-click获取关键字段 and log explorer stack mode displaying key fields, instead displaying based on display items.
Fixed critical level errors in monitor execution logs.
Fixed DQL subquery ineffectiveness.
Fixed incorrect condition filling due to unescaped commas in variable values when clicking leaderboard view and selecting apply to view variables.

May 14, 2025¶

Brand LOGO Change¶

Guance adopts an innovative visual image, demonstrating its iterative belief and technological strength in the AI wave.

From this moment on, Obsy will accompany you to洞察 the data world,驾驭 system complexity, and walk at the forefront of the AI era.

External API¶

Dashboards: Supports global import of platform-level dashboards via identifier ID. This operation only updates dashboard configuration,不影响 the binding relationship of this view in each workspace.

Feature Updates¶

Security Monitoring¶

Added SIEM function module: Real-time analysis of logs and event data from various enterprise systems (such as servers, applications, network devices), automatically discovering potential threats, helping teams quickly locate anomalies, and fully leveraging the role of a security monitoring hub.

Monitoring¶

Alert Policies:
Association: Supports selecting SIEM detection rules;
Notification Configuration: Added "Security Monitoring" type level.
Monitors: Due to adjustments in the "Infrastructure Change Detection Monitor" data writing, old monitors need to be re-saved to avoid affecting generated event display.

Event Center¶

The "Events" function module renamed to "Event Center", interaction page optimized.
Added Event Center explorer, supports centralized viewing of all event data.
Event recovery新增 secondary confirmation operation to avoid accidental touch scenarios.

Incident¶

After adding schedules in Incident > Calendar, when creating new Issues or associating responsible persons, you can choose "On-duty Personnel" or "Notification Targets". The system will automatically assign Issue responsible persons based on the day's on-duty personnel and notify them.

Scenarios¶

Dashboard: Added Obsy AI Intelligent Analysis function.
Time widget logic optimized: When selecting "Relative Time", even if the user sets the refresh frequency to "off", modifying query conditions (e.g., adding/removing conditions using quick filters in the explorer, adjusting filter conditions in the search box, or switching view variables in dashboards/notes) will still trigger a re-query.

Management¶

Data Forwarding: For AWS S3, Huawei Cloud OBS, Alibaba Cloud OSS these three archive types, added Encrypted Storage selection switch. When enabled, the system encrypts outwardly forwarded data.
Audit Events: Added workspace field marking.

Infrastructure¶

Host, container, Pod, etc. detail page built-in pages:
Time Widget: Added Pin function;
Integration Operation Status: "View Monitoring View" page optimized;
Associated views新增 current event occurrence timeline display.
Host Explorer: Default query time adjusted to "Last 2 hours".

Logs¶

Optimized log multi-index matching hit logic, avoiding computational waste caused by逐条 matching index rules.

Deployment Plan Updates¶

Supports creating, deleting, and modifying groups via External API, and adding or removing users in groups;
Workspace Management > Data Storage Policy: Added network storage policy configuration, supports selecting fixed 2-day storage duration or custom保存 duration.
Added Profile data reporting dynamic parameter switch. Users can dynamically enable or disable Profile data reporting based on platform performance.
Added API Key configuration and AI function access link configuration switches, including the following scenarios:
Monitors > Alert Policies > AI Alert Aggregation
AI Write Pipeline entry
Scenarios > Chart Components > AI Chart Title and Description Generation
Log Explorer > Log details with error status, Logs > Error Tracking > AI Error Analysis entry
RUM > Error Explorer > AI Error Analysis entry
APM > Trace Explorer > Trace details with error status, APM > Error Tracking > Trace Details > AI Error Analysis entry
When editing alert policies, supports custom repeat alert time. Custom input mute time format example: 20m (20 minutes), 2h (2 hours), 1d (1 day).

New Integrations¶

Added host-class monitors;
Added MySQL monitors;
Updated CouchDB views and monitors;
Added ActiveMQ integration;
Added AWS SNS integration;
Added Volcengine ESCloud integration;
Added English versions of Huawei Cloud API, Huawei Cloud CCE monitors and dashboards;
Added Volcengine RabbitMQ integration;
Modified AWS MSK integration metrics.

Bug Fixes¶

Fixed 413 errors when uploading SourceMap via Vite packaging, SourceMap download errors, and Guance inability to display source code error locations.
Fixed inability to display multiple spaces RUM data normally during cross-workspace queries.
Fixed inaccurate namespace and service data display after selecting self-built log indices in the log explorer.
Fixed messy page styles when creating multi-step dial tests.
Fixed inability to normally display namespace tag data in the quick filter column under multi-index conditions in the log explorer.
Fixed false alert误报 on the Guangzhou site.
Fixed异常 when viewing host details using hosted Doris in version 0409.
Fixed abnormal external data source query data.
Fixed ghosting display when expanding single logs in log clustering mode.
Fixed still displaying recovered events on the event statistics page.
Fixed errors triggered by clicking the "Batch" button in unrecovered event list display mode.
Fixed persistent inability to query historical logs on S3 using the "Data Forwarding" function.
Fixed FUNC alert message sending logs from version 1.109.196 Volcengine base being reported to business spaces.

April 23, 2025¶

Breaking Change¶

Dial Test Data Storage Adjustment: Separated dial test data from the original log index, independently stored in tables, and added exclusive query Namespace D. This adjustment does not affect historical dial test data. If custom charts were previously used to query dial test data, please update the query configuration promptly. Official dial test templates (e.g., HTTP, ICMP, TCP, WebSocket, Multi-step Dial Test) and related monitors, metric rules have been adapted to this change.

External API¶

Supports新增, modifying listing specified workspace role lists, member lists interfaces;
Optimized interface names and descriptions in OpenAP, External API.

Feature Updates¶

RUM¶

SDK supports dynamic updates via Environment Variables: Achieve dynamic configuration updates by pre-defining variable parameters and introducing variables through environment variables within the application.
Applications > New Application: The left-side code example updates in real-time based on user custom parameter configuration, the system caches the latest parameter configuration, no need to re-enter each time.
Error Explorer: Added support for Obsy AI Error Analysis. Intelligently parses single RUM error logs, automatically extracts key information, and obtains preliminary solutions and suggestions through integrated online search.

Infrastructure¶

Cross-Workspace Authorization: If authorized workspaces exist and the authorized data type is "Basic Objects", you can directly switch workspaces on the infrastructure various object type explorer pages to achieve cross-workspace object data viewing.
Added "Change Events" tab page, supports displaying change details based on Deployments, Daemonsets, Services, Cron Jobs, Statefulsets five object type data,便于 users回溯 change records.

Monitoring¶

Monitors > Detection Rules: Added "Infrastructure Change Detection". By continuously monitoring Kubernetes resource configuration changes (e.g., YAML file modifications), combined with version tracking and intelligent analysis, achieves compliance checking, fault root cause positioning, and resource usage optimization effects, ultimately ensuring cluster security, stability, and resource efficiency.
Detection Rules > Event Content > Custom Event Notification Template: Added time auto-conversion Template Function {{ df_fault_duration | to_date_range_human }}.

Events¶

Added Change Events:
View all event data records triggered based on Infrastructure Change Detection rules.
All Event Details > History Trend: When the detection object has change events, the time series chart displays the change event time nodes and supports clicking to jump to the change details page.
Event Explorer:
Added "Only Show Unassociated Issues Button", clicking quickly filters out associated events;
Added "Display Items": Supports display preference switching (including list, standard, extended三种方式);

Management¶

Data Forwarding > Data Viewing: Search Component Optimized.

Chart Export¶

Charts support exporting to newly created dashboards or newly created notes synchronously, applicable to: 1. Metric Analysis 2. Explorer > List/Dimension Analysis

Explorers¶

Adjusted interaction for basic "List" view and "Dimension Analysis" chart mode;
Dimension Analysis: Added grouped table chart.

Integrations¶

In the console left menu, the "Cloud Account Management" function access path has been moved from the original Management interface to the Integrations directory.

Deployment Plan Updates¶

Supports adjusting Login Method (Local Account, LDAP Account, OIDC) order.

New Integrations¶

Added apollo;
Updated Argocd integration documentation, adjusted to kubernetesprometheus for collecting metric information;
Updated ingress nginx documentation views monitors;
Neo4j updated dashboard & added monitor;
Added kubelet integration;
Added Azure Virtual Network Gateway integration;
Updated TIDB integration documentation, views and monitors;
Alibaba Cloud 30 components added Automata management.

Bug Fixes¶

Fixed incorrect member count display on the member management page.
Fixed untranslated log view time fields after switching to English interface.
Fixed abnormal label synchronization display after switching clusters.
Fixed stutter count affecting user count display异常 in RUM > Analysis Board > Mobile > Error Analysis page when mobile SDK was not integrated.
Fixed timing judgment issues on the log explorer > details page.

April 9, 2025¶

Open API Updates¶

Supports creating, editing, deleting multi-step dial test tasks;
Supports configuring workspace quantity上限限制.

Feature Updates¶

Scenarios¶

Chart Optimization
Bar Chart:
- Adjusted alias position, supports listing all metrics and groups;
- Added X-axis configuration item.
Grouped Table Chart:
- Supports sorting based on group selection;
- Added 200 option in return count dropdown, maximum manual input adjusted to 1,000.
Time Series Chart > Line Chart: Added line "Style" setting, including linear, smooth, step-before, step-after.
Snapshots: Added permission hints for dashboard snapshot sharing with configured permissions.
Explorer, Dashboard > Time Widget: Added "Last 1 minute", "Last 5 minutes",默认 selects the latter.

Management¶

Cross-Workspace Authorization: Supports cross-site data authorization, enabling extended data sharing.
Data Forwarding:
Data type新增 "Audit Events".
Adjusted explorer query time logic:
- Changed to daily query, does not support cross-day queries;
- When viewing forwarded data, the system automatically queries and continuously loads until complete display, no need for user manual clicks;
- Management > Workspace Settings > Advanced Settings: Added "Data Forwarding Query Duration" configuration.
Data Access, Pipeline, Blacklist permissions split, "Management" permission adjusted to: "Create, Edit" and "Delete".

Monitoring¶

System Notifications: Added associated log jump links, allowing jumps to the log explorer and filtering out logs where notification object sending failed.
Monitors:
New from Template > Official Template Library: Added search function;
Configuration Page > Event Content: Updated note. Only when关联 incident is enabled, will the @ member configuration生效 and send the event content here to the specified members.
Threshold Detection: Added Recovery Condition switch, configure recovery conditions and severity. When the query result has multiple values, any value meeting the trigger condition generates a recovery event.

Incident¶

The "Source" in Issue email notifications changed to a hyperlink, users can access directly after clicking.

RUM¶

SourceMap: - Page interaction adjusted, changed from popup to separate page; - List page新增 search, export functions.

Obsy AI Agent¶

Optimized Generate Chart function: Caches metric management data locally via Func, generated DQL is closer to semantic description.

Obsy AI Error Analysis ¶

Added context support for root cause analysis, helping users understand the context of error occurrence faster and more comprehensively, improving diagnosis efficiency.

Integrations > Extensions¶

DataFlux Func Hosted Edition/ RUM Headless: Application deduction新增 email reminders and system notifications.

Infrastructure¶

Hosts > Details Page: Disk capacity statistics distinguish between local disks and remote disks for display.

Deployment Plan Updates¶

Template Management: - Page interaction optimized, includes adding "Template Type" to display columns, etc.; - Top filter items新增 "Template Type"; - Supports batch exporting templates; - When importing templates, supports previewing template details; - Built-in views, Pipeline, monitor template upload supports同名覆盖 logic.

New Integrations¶

Added GCP Compute Engine;
Added Azure Storage;
Added Azure redis cache;
Added Azure kubernetes;
Added Azure Postgresql;
Added Alibaba Cloud Rds MYSQL Automata integration;
Added Druid integration;
Updated Trino;
Adjusted AWS/Alibaba Cloud integration documentation regarding Automata: Added Hosted Edition Activation Script step.

Bug Fixes¶

Fixed Open API获取 unrecovered events not matching reality.
Fixed errors when searching in the event explorer.
Fixed abnormal external data source data queries.
Fixed incident email notification related issues.
Fixed slow incident > analysis board loading.
Fixed incorrect bar chart display colors.
Fixed no AK提示 for external events.
Fixed status distribution chart, log volume not displaying correctly.

March 26, 2025¶

Feature Updates¶

Events¶

Unrecovered Events: Time widget defaults to auto-refresh, manual event recovery新增 frontend page prompt after successful recovery.
Event Details: Alert notification page display optimized.
Event, intelligent monitoring explorer export appended "Notified or not" display.

Management¶

Data Forwarding: Added permission configuration. By setting view permissions for forwarded data, data security is effectively improved.

Monitoring¶

Infrastructure存活 Detection V2: Added configurable detection interval.
Alert Policies: Notification matching logic adjusted, improving execution efficiency in scenarios where events hit multiple policies and complex alert policies.
Notification Target Management: Added Slack and Teams two notification targets to meet the needs of more global users.

Scenarios¶

View Variables: Supports external data source queries.
Charts:
Added "Monitor Overview" chart, integrating monitor list, displaying latest status, achieving instant monitoring and异常态势感知.
Command Palette: Display optimized;
Chart Query: In DQL queries, when using "wildcard" or "regex",新增 query suggestion hints.
Explorer, Dashboard > Time Widget: Added "Last 1 minute" and "Last 5 minutes" options.

Billing¶

Added Monthly Bill module, intuitively displaying monthly total consumption.
Added export bill function.

RUM¶

Application List > New Application: Added "Compressed Upload" and "Custom Hosted Address" parameter configurations.

Synthetic Tests¶

Added "South Africa" and "Hong Kong, China" as dial test nodes, further expanding global coverage.

Logs¶

Log Explorer
Quick filter operation items adjusted;
List tokenization logic optimized;
JSON format data新增 "JSON Search";
Index > Key Fields: Added "One-Click Extract".

New Integrations¶

Added host-class monitors;
Added MySQL monitors;
Updated CouchDB views and monitors;
Added ActiveMQ integration;
Added AWS SNS integration;
Added Volcengine ESCloud integration;
Added English versions of Huawei Cloud API, Huawei Cloud CCE monitors and dashboards;
Added Volcengine RabbitMQ integration;
Modified AWS MSK integration metrics.

Bug Fixes¶

Fixed low utilization display on the log page.
Fixed Service Map metric unit display issues.
Fixed inability to select units for multiple columns in table charts.
Fixed errors when exporting dashboard > log stream graph to CSV with counts other than 1,000.
Fixed inconsistency between most popular page P75 and DQL query results.
Fixed time框 still showing today after clicking << button in time filter.
Fixed menu management not meeting expectations.
Fixed abnormal space ID search filtering in management console.
Fixed lost test samples in Pipeline interface.
Fixed overly long configuration migration export function耗时.
Fixed大量空项 in quick filter labels on the event details interface after upgrade.
Fixed duplicate detection library lists in monitor official template library, and inability to search other monitors once a detection library is selected.

March 12, 2025¶

Breaking Changes¶

Event df_alert_info field definition adjusted, added alert policy unmatched reason description, still needs filtering via isIgnored to obtain actual externally sent notification targets.

Feature Updates¶

Incident¶

Added Incident Management entry. Through this entry, the currently logged-in user can view and manage the incident status of all workspaces they have joined.
Optimized incident page channel list display, improving query efficiency when there are too many channels.

Management¶

Cloud Account Management: Added Alibaba Cloud cloud account authorization type.
API Key Management: Added permission control for API Keys, supports adding role authorization. Through role authorization, API Keys only have operation permissions within the role scope, effectively reducing security risks.
Data Forwarding: Default interaction changed to not select rules.

Obsy AI Error Analysis¶

The following detail pages added Obsy AI Error Analysis capability: - error logs - APM > Traces/Error Tracking

Scenarios¶

Scheduled Reports:
Added Webhook sending as a notification method;
Supports sharing dashboard images to WeCom/DingTalk.
Time Series Chart: After selecting area chart as the chart type, added Stack Mode style,便于 observing the cumulative effect of overall data.

APM¶

Traces: Supports batch export of lists in JSONL format.

RUM¶

User Insight > Funnel Analysis: Supports session replay functionality for the queried Session list.

Logs¶

Explorer:
Log Explorer > Index quick filter display effect optimized in search bar;
Log Details > Extended Fields: Added "Perform Dimension Analysis" mode;
Index: Supports setting exclusive Key Fields under the index dimension.

Left-side time range selection and right-side refresh frequency are independent. Only two situations affect refresh frequency: - Selected time range exceeds 1h - Selected time is absolute time

Infrastructure¶

Hosts: Explorer supports adjusting time range.

Pipeline¶

Configuration page display optimized;
Pipeline processing type新增 "Events";
Test samples support obtaining JSON format.
Filter Conditions > Synthetic Tests: Supports selecting multi-step dial tests.

Deployment Plan Updates¶

Template Management: Supports uploading all explorer templates.

New Integrations¶

Added azure_load_balancer;
Rewrote K8S server api;
Updated Gitlab CI;
Translated Volcengine related views;
Translated AWS related views.

Bug Fixes¶

Fixed no response when exporting log stream graph to CSV.
Fixed no data after adding relevant filters to time series charts when the view variable for ddtrace collected JVM metrics was the runtime-id field.
Fixed custom gradient interval color scale interface display issues.
Fixed empty display when re-editing time series chart DQL queries after saving with filter condition >0.
Fixed abnormal APM > Infrastructure table chart display.
Fixed frontend forwarding rule incompatibility after setting data forwarding storage duration to 1,800 days in management console.
Fixed "kodo service API request error: Service Unavailable" when executing show_object_field(HOST) in quick query.
Fixed bugs in shortcuts.
Fixed no data for RUM session and view, while other resource and action, etc. had data.
Fixed immediate validation of required fields when creating multi-step dial test request steps.
Fixed filter conditions not生效 when setting role authorization in data access.

February 27, 2025¶

Open API Updates¶

Metrics: Added measurement and tag information acquisition.

Feature Updates¶

Synthetic Tests¶

HTTP Dial Test: Supports Script Mode. By writing Pipeline scripts, flexibly customize the judgment conditions and data processing logic of dial test tasks.
Added Multi-step Dial Test: Allows creating tests using multiple API connection response data, and linking multiple task requests through local variable value passing.

Scenarios¶

Dashboard > Visibility: Added "Custom" configuration, can configure "Operate", "View" permissions for this dashboard. Simultaneously, this configuration新增 "All Members" option.
Charts:
Added ability to automatically generate chart titles and descriptions based on AI;
Log stream graph新增 "Rule Mapping" function;
Table chart display columns optimized;
Grouped table chart: Expression results support sorting;
Time series chart, pie chart, and multiple other charts support data export to CSV files.

Metrics¶

Metric Analysis > Table Chart: When query results exceed 2,000条, three modes新增 "Query Result Count" display.
Metric Management: Supports one-click jump to metric analysis.
Generate Metrics: Supports import creation method, supports batch export.

Infrastructure¶

Container/ Pod Explorer: Object data新增 cpu_usage_by_limit, cpu_usage_by_request, mem_used_percent_base_limit, mem_used_percent_base_request 4 new fields.

APM¶

Profiling > Flame Graph interaction optimized: Selecting a single search method name可以直接 focus定位.
ServiceMap interaction optimized: In the upstream/downstream pages, supports searching for nodes in the current canvas.

New Integrations¶

Added AWS cloud bill;
Added Kube Scheduler;
Added MQTT;
Rewrote APISIX;
Updated tidb English documentation and views;
Updated Zookeeper views, added integration icons;
Fixed English translation for some component mainfest.yaml.

Bug Fixes¶

Fixed errors when clicking APM > trace details tab page;
Fixed incorrect @member in Incident > Issue replies;
Fixed incorrect temperature units in charts.

February 19, 2025¶

Breaking Changes¶

Event df_meta will no longer retain alert_info related information records. Users who previously relied on this information to obtain notification targets should switch to using the newly added df_alert_info (event alert notification), df_is_silent (whether muted), df_sent_target_types (event notification target types) 3 fields to complete corresponding functions.

Potentially affected functional scenarios: 1. Custom usage scenarios integrating with external systems via Open API获取 events 2. Custom usage scenarios forwarding events to external systems via Webhook notification targets

Feature Updates¶

PromQL Query¶

Added query type: Instant Query, i.e., query for a single time point.

Monitoring¶

Monitor configuration page: 1. Added not between option in trigger condition logic matching; 2. Supports directly modifying monitor status ("Enable" or "Disable").

APM¶

Traces: Details page新增 Service Context tab page.

Events¶

Event Details Page: Added support for binding Built-in Views;

Incident¶

Issue新增 working, closed statuses;
For open status timeout and unspecified responsible person and processing timeout situations, Issue Escalation新增 repeat notification configuration;
Adjusted Issue system comments, channel notification UI display;
Analysis Board: Added time widget.

Scenarios¶

Chart Link: Added "View Host Monitoring View",默认 off.
Explorer: Supports deleting fixed name column, users can customize list display.
Cloud Bill Analysis View: Supports viewing bill details.

Management¶

Role Management: Session Replay viewing, Audit Events新增 custom add view permission capability.

Metrics¶

Generate Metrics: Metric name input no longer supports using - hyphen.

Integrations¶

Integration cards新增 description information.

Deployment Plan Updates¶

Template Management: Supports uploading infrastructure explorer templates;
Index Configuration: Deprecated "Backup Logs" item; Can configure corresponding storage policy at "Edit Workspace > Data Storage Policy > Data Forwarding-Default Storage".

New Integrations¶

Added Milvus;
Added Volcengine Public IP;
Added opentelemetry-python;
Added openLIT integration;
Updated k8s\es\mongodb\rabbitmq\oracle\coredns\sqlserver Chinese/English monitors & views.

Bug Fixes¶

Fixed result abnormalities caused by special characters in AI aggregation notification messages;
Fixed Servicemap deployment version adaptation issues;
Fixed inability to configure hidden view variables in combined charts;
Fixed messy display of "Unrecovered Issue List" in Incident > analysis board;
Fixed inconsistency between most popular page P75 results and DQL query results in RUM analysis board;
Fixed RUM > explorer search box异常;
Fixed only partial effectiveness of the same field in boards when using resource catalog for field mapping in Scenarios > object mapping;
Fixed UI display issues in Monitors > event content;
Fixed unmet expectations for quick filter results in event explorer unrecovered events.

January 16, 2025¶

Feature Updates¶

RUM¶

Added Funnel Analysis function: Users can create funnels by defining conversion steps, view data conversion and conduct in-depth analysis;
User Insight module integration: Added User Insight module, integrating heatmap and funnel analysis into this module, providing more comprehensive user behavior analysis tools;
Added mobile端 SourceMap restoration: Android and iOS applications support uploading SourceMap files on the page and viewing restored data in the error explorer.

APM¶

When adding services in APM, added installation guidance method for Host Auto-injection, simplifying the installation process.

Integrations¶

DataKit (Data Collection Tool): DataKit installation page added Docker method installation guidance, providing more diverse installation options;
External Data Source Optimization: Added query specification hints when querying SLS data sources, helping users perform data queries more accurately.

Scenarios¶

Combined Chart optimization: Combined charts新增 view variable configuration, supports selecting view variables from the current dashboard to apply to this combined chart, helping with more flexible filtering and data analysis.

Monitoring¶

Mutation Detection Monitor: Added support for week-over-week, month-over-week comparison of query periods.

Obsy AI Agent¶

Added DataFlux Func related knowledge base.

Pipeline¶

Auto-generate Pipeline optimization: Supports interactively obtaining Pipeline parsing simultaneously in structured plus natural language ways.

Bug Fixes¶

Fixed display issues in log stack mode;
Fixed misaligned function input box in log detection monitors;
Fixed incorrect metric operations;
Fixed Volcengine not supporting having statements;
Fixed errors when selecting "Request Error Rate" and "Average Requests Per Second" two metrics in APM metric detection;
Fixed not in statements not生效 in Volcengine base;
Fixed large event list returned data affecting page loading speed;
Fixed unmet expectations for one-click event recovery on Hangzhou site.

January 8, 2025¶

Open API Updates¶

Field Management: Supports获取 field management list, supports add/modify/delete field management.
Synthetic Tests: Supports modify dial test tasks.
Incident > Schedule: Supports获取 schedule list, supports create/modify/delete schedule.
Incident > Configuration Management: Supports获取 notification policy list, supports add/modify/delete notification policy; supports获取 Issue discovery list, supports create/modify/enable/disable/delete Issue discovery configuration.

Feature Updates¶

Logs¶

Log Index Optimization:
When accessing built-in log views, log context tab pages, will默认 select the current log's index, default index respectively. Both tab pages support multi-select of indices. Simultaneously, when cross-workspace query is enabled and authorized workspaces are selected in the belonging menu, supports directly querying the corresponding workspace's index data here. Ultimately helps users completely view associated log data on one page, optimizing log query interaction.
When listing log indices, besides default pinned display, other log indices are listed in A-Z order.
Log Explorer新增 stacked View Mode: In stack mode, fields will be integrated into the same column, and these fields are presented in rows within the cell. Log information display is more compact and clear,便于 users quickly browse and analyze.
Log Pipeline Optimization: Log Pipeline test samples adjusted to obtain all fields of the log, and need to be filled in line protocol format. Simultaneously, logs manually entered by users must also follow format requirements.

Scenarios¶

Table Chart optimization:
Multi-metric query sorting support: When using one DQL for multi-metric queries, table charts now support sorting.
Table pagination selection: Added table pagination selection function, users can choose appropriate pagination size based on data volume and viewing needs.
Combined Charts: Supports adjusting chart order.
Chart Optimization: Adjusted DQL query component function order, simultaneously specially emphasized Rollup function usage scenarios, helping users better utilize Rollup function for data aggregation and analysis.

Management¶

Events support configuring Data Forwarding: Supports configuring data forwarding rules for event type, saving event data meeting filter conditions to Guance's object storage and forwarding to external storage, providing flexible event data management capability.
Workspace新增 DataKit Environment Variables: Workspace supports managing DataKit environment variables, users can easily configure and update environment variables, achieving remote synchronous update of DataKit collection configuration.
Query Audit Events optimization: Added multiple fields for recording query information, simultaneously supplemented query time range in event content,便于 tracking and analyzing query behavior.

Pipeline¶

Auto-generate Pipeline optimization: Changed prompt appearance method, optimized product experience.

Obsy AI Agent¶

Obsy AI Agent新增 Generate Chart: Generate chart function automatically analyzes user-input text data based on large models, intelligently generates suitable charts, solving problems like繁琐 manual chart creation, chart selection difficulty.

Monitoring¶

Alert Policies: Configuring notification rules by member supports appending names for purpose description.

Deployment Plan Updates¶

Management Console > Workspace Menu Optimization:
Workspace list新增 main storage engine, business two filter items, supports convenient workspace filtering;
Optimized workspace list page number return logic. When modifying/deleting a workspace, or modifying workspace data reporting limits, will stay on the current page, optimizing query experience.
Deployment Edition新增 parameter: alertPolicyFixedNotifyTypes, supports configuring whether the "Email" notification method is displayed in alert policies Configuration Reference.

New Integrations¶

Added AWS Gateway Classic ELB;
Added Volcengine TOS Object Storage;
Modified AWS Classic collector name;
Added MinIO V3 integration;
Updated elasticsearch, solr, nacos, influxdb_v2, mongodb integrations (views, documentation, monitors);
Updated kubernetes monitoring views.

Bug Fixes¶

Fixed event data cross-space authorization not生效;
Fixed inability to query data after log associated trace jump to trace explorer carrying trace_id;
Fixed inability to perform numerical filling for view expression queries;
Fixed lack of operation audit records when changing alert policies for external event detection monitors;
Fixed unadjustable column width in event display list.