Changelog

Note:

This page is currently under development. If you have any feedback or encounter issues, please contact our team.

3.8

Server

1. Enhanced the Summary Report with problem-specific tags, along with clearer Problem and Recommendation statements, enabling faster issue identification and more effective troubleshooting.

2. JFR Report Enhancements:

   • Added an Environment Report to provide visibility into environment variables, system properties, and related configuration details.
   • Added a Memory Report for deeper insights into JVM memory usage and allocation activity.
   • Enhanced error and warning messages to provide clearer guidance for troubleshooting.
   • Added an Environment Comparison section to easily identify differences in configuration settings.

3.7

Server

• JFR Bug Fixes: Resolved various issues related to Java Flight Recorder (JFR) file processing and analysis.

• Async-Profiler JFR Support: Added support for parsing JFR files generated using the async-profiler command.

• Optimized Bundle Uploads: Improved bundle upload performance by splitting large files into smaller chunks for more efficient processing.

• Extended JFR File Format Support: Added support for uploading JFR files in .gz, .tar.gz, and .tgz formats.

• Fixed Regex CPU Spike Issue: Resolved an issue that could cause excessive CPU utilization during regex processing.

yc-360 Script

• Go Dependency Security Updates: Updated golang.org/x/text to v0.39.0 to address vulnerability GO-2026-5970.

• Compression Library Security Update: Updated github.com/klauspost/compress to v1.18.7 to address vulnerability GO-2026-5841

3.5

Server

• Added support for Java Flight Recordings (JFR). You can now upload .jfr files directly from the Bundle Upload page or via the Bundle Upload drop-down in the top-right corner of the Calendar Dashboard.

• Introduced Report Menu Indicators to highlight report status. A green dot indicates that content is available with no errors, while a red dot indicates that the report contains errors.

• Enhanced the LLM Incident JSON to include JFR data, including File I/O, Exception, and Method Profiling information.

• Improved thread dump upload performance by uploading files in chunks, resulting in faster and more reliable uploads.

• Optimized thread dump analysis to improve processing speed and overall performance.

3.4

Server

• Optimized fastThread to process large thread dumps (up to 32 GB) faster while significantly reducing memory consumption.

• Fixed multiple issues in the M3 Dashboard to improve stability and reliability.

• Resolved the upload issue affecting compressed thread dump files.

• Enhanced email notifications with a redesigned template and additional information fields. As part of this enhancement, the AI Overview codebase has been refactored.

• Added direct hyperlinks in the AI Overview section, allowing users to navigate directly to the relevant report pages.

• Introduced fastThread AI with a new Deterministic AI tab. This feature provides AI-powered thread dump analysis and an interactive chat interface for answering thread-related questions, enabling faster identification and resolution of thread issues.

• Fixed several issues in GCeasy AI to improve functionality and user experience.

• Addressed multiple bugs in AppLog.

• Added support for .zst and .zstd compressed formats in the Bundle Upload feature.

• Implemented a fallback upload mechanism in GCeasy and fastThread to ensure analysis continues even if AWS S3 uploads fail from the Cool Tools page.

• Fixed issues in HeapHero Cool Tools web upload functionality to improve performance and reliability.

yc-360 Script

• Replaced ZIP compression with Zstandard (zst/zstd) for only-capture results and heap dump files, providing improved compression efficiency.

• Upgraded the Go toolchain to v1.26.4 to address reported standard library vulnerabilities.

• Updated vulnerable Go dependencies and rebuilt binaries using the latest patched Go runtime.

• Fixed the temporary heap dump file cleanup sequence to ensure proper resource management.

3.3

Server

• Optimized the repeating statements detection logic to improve processing performance and reduce analysis time.

• Introduced a Scorer mechanism in Error Analysis to minimize false-positive error detections.

• Added a new feature: RCA Executive Summary. An AI-generated summary is now displayed under the "AI Overview" section on the yCrash RCA page. When an artifact bundle is uploaded, AI Overview automatically generates a concise incident summary by analyzing captured artifacts such as GC logs, thread dumps, heap substitutes, system metrics, network analysis reports, device errors, and application logs. Users can also select "Dive Deeper in AI Mode" to launch yCrash Buddy and ask follow-up questions related to the incident.

• Added the Swagger API Documentation page for easier API exploration and integration. Additional details are available here (opens new window)

• Fixed issues related to .NET log analysis.

• Introduced AppLog Comparison, enabling comparison of multiple application logs within the Compare Report section.

• Added GCeasy AI support. GCeasy now includes a Deterministic AI tab that delivers AI-powered GC log analysis and provides answers to GC-related questions through an interactive chat interface, helping users identify and resolve GC issues more efficiently.

yc-360 Script

• Added support for .NET applications in M3 mode.

• Introduced automatic AppRuntime detection, removing the requirement for manual appRuntime configuration.

• Enhanced error messages across the application to improve troubleshooting and user experience.

• Added a .NET helper utility to capture logs from 32-bit Windows systems.

• Upgraded the Go toolchain to v1.26.3 to address reported standard library vulnerabilities.

• Updated vulnerable Go dependencies and rebuilt binaries using the latest patched Go runtime.

• Resolved compatibility issues in the golangci/golangci-lint-action CI workflow caused by the Go version upgrade

3.2

Server

• Fixed an issue with heap dump uploads on the individual upload page.

• Introduced support for mapping custom documentation pages instead of loading content from the default documentation site.

• Added visibility of discarded incidents in the M3 actions timeline.

• Resolved multiple bugs in M3.

• Implemented memory size validation for large application log files.

• Addressed security vulnerabilities in the webapp-runner and jackson-core JAR files.

yc-360 Script

Upgraded to Go 1.25.8 and updated related dependencies (x/net/sys/term/text) to address known security vulnerabilities.

3.1.5.1

Server

Introduced support for invoking yCrash REST APIs without a static API key. For more details click here (opens new window)

yc-360 Script

Added support for OpenJ9 in heap substitute data collection.

3.1.5

Server

• Redesigned and reimplemented repeating statement detection with new logic to accurately identify and print patterns across application log lines.

• Added a new View AI Solution feature to the RCA Report, AppLog Report, and Error Sub-report pages, allowing users to generate AI-driven solutions for each AppLog error. For more details about this feature click here (opens new window)

• To reduce false positive incidents reported by M3, two new system properties have been added: monitorGC (opens new window) and monitorThread (opens new window). By default, M3 now monitors only macrometrics. To enable micrometrics monitoring, configure -DmonitorGC=true and -DmonitorThread=true in the launch script.

• Enhanced AppLog parsing logic to better support diverse unstructured log formats.

• Improved performance for large application logs and RCA Reports.

• General improvements and bug fixes.

yc-360 Script

Minor improvements and bug fixes

3.1.4

Server

• Introduced .NET reports for GC, thread dumps, and heap substitute, enabling these reports to be viewed in the yCrash Server dashboard.

• Updated backend components to remediate vulnerabilities identified in third-party libraries.

• Upgraded client-side libraries to address reported security vulnerabilities.

• Improved HeapHero upload logic to enhance overall performance.

• Resolved multiple front-end issues in HeapHero.

• Prevented duplicate JVM entries from being created when the target application restarts.

• Replaced the Sunburst chart with a TreeMap visualization in the HeapHero Large Objects section.

• Implemented Acute Memory Leak detection logic in GCeasy for supported log files.

• Updated the Repeating Statements section in App Logs to display only the top 100 statements/patterns.

• Enhanced the auto-delete mechanism to clean up old logs and remove M3 samples from the database based on the configured retention policy.

• Enhanced AppLog support for large files, improving stability and load performance.

• Improved AppLog sub-report page usability and performance by migrating tables to AG Grid.

• Fixed minor defects in M3.

• Upgraded the UI/UX for hs_err_pid reports.

yc-360 Script

• Introduced new flags to support capturing .NET artifacts from target applications. These flags allow the yc-360 script to capture artifacts in onlyCapture and onDemand modes. For more detail about this feature click here (opens new window)

• yCrash .Net Command: Introduced a new .NET command to capture artifacts such as GC, thread dumps, and heap substitute.

• Introduced a new argument minimalTouch to skip the CPU-intensive capture.

• Improved the performance of file-based log capture, with optimizations specifically for large files.

• Added a warning log when the processTokens option is used without enabling M3 mode.

• Enhanced AppLogs behavior in on-demand mode to log a warning and capture all logs specified under the -appLogs argument instead of skipping them.

3.1.3.1

Server

Upgraded tika-parsers library to fix the vulnerability issues.

3.1.3

Server

1. Micrometrics M3 Integration: Added response time monitoring as part of Micrometrics (M3) tracking.

2. Discard AppLog Feature: Introduced automated filtering of irrelevant log statements (e.g., database names, NullPointerException) using a configurable discard file — ensuring cleaner and more focused analytics. Learn more (opens new window)

3. Statement-Level HTTP Status Count: Implemented statement-level tracking for HTTP status codes to provide finer-grained visibility.

4. Enhanced Error Visibility: Full error details are now displayed in the AppLog Observation and Error pages for improved troubleshooting.

5. NMT Report Fixes: Resolved known defects in the NMT reporting module.

yc-360 Script

Response Time Monitoring: To monitor your application's response time, ensure that access logs are enabled. Learn more (opens new window)

3.1.2

Server

1. Allocation Affected Threads: A new section has been added to display threads impacted by allocation stalls. This enhancement provides better visibility into allocation slowdowns and their effect on overall application performance.

2. Log Growth Optimization: Optimized the application’s log management process to reduce unnecessary log file growth, improving both storage efficiency and performance.

3. App Log Download Option: Introduced a download option for application log files within the App Log Report, enabling easier access and sharing for analysis.

4. Statement-Level Logging in Application Log: Added a new feature to categorize log entries by level (Info, Error, Warning), enabling more detailed log analysis and streamlined troubleshooting.

5. PostgreSQL DDL Script Management: PostgreSQL schema changes are now externalized into ddl_postgres_ee.txt, allowing for automated or manual database initialization using the -DcreateDBManually property. Each release's schema updates are versioned for easy tracking and auditing. For more details, refer to this documentation here (opens new window).

6. Bug Fixes:

   • Fixed a UI distortion issue that occurred after installing a new version due to browser cache.

   • Resolved an issue where SSO logout was not functioning correctly with ADFS by refining the logout flow.

   • Normalized the CPU% values in the Processes Report based on CPU count for accurate reporting.

   • Addressed multiple issues related to log accuracy, formatting, and level-based categorization.

yc-360 Script

1. Fixed vulnerability in the oauth2 package

2. Upgraded Go module to version v1.23.12

3. Minor bug fixes and performance improvements

3.1.1

yc-360 Script

Added cmdTimeout to control how long yc-360 waits for external tools to finish capturing thread dumps or logs. For more information refer here (opens new window)

3.1

Server

1. Heap Dump Analysis

• Added discard options to efficiently analyze large heap dump files.For more information refer here (opens new window)

• Introduced a "Class Hierarchy" view in the Actual Data section of the Histogram.

• Optimized performance when loading child nodes of large datasets in Largest Objects, Incoming References, and Outgoing References pages.

• Set the “Static” tab as the default view in the Histogram’s Actual Data table.

• Resolved hyperlink issues in problem statements that link to thread stack traces.

2. Infrastructure & Configuration

• Fixed a database connection pool leak.

• Removed system properties from saml.xml to streamline SAML configuration.For more information refer here (opens new window)

3. Customization

• Added support for configuring a custom company logo.For more information refer here (opens new window)

4. Thread Analysis Enhancements

• Introduced a new "Platform vs Virtual Threads" section in the thread report for better comparison and analysis of traditional vs virtual threads.

5. Applog Report Improvements

• Migrated the table section of the Applog report page from DataTables to AG Grid for improved performance and functionality.

• Fixed several bugs to enhance the stability and overall user experience of the Applog report page.

yc-360 Script:

• Bug fixes and improvements for Extended Data capture

3.0.9.4

Server

• Introduced support for m3SamplesRetentionPeriod to configure the retention period of M3 data in both local and remote storage. Cleanup is triggered by existing schedules: runJobAt (remote) and runLocalTempJobAt (local). If not set, the default retention is 1 day. More details (opens new window)

• Added support for custom scripts to collect additional artifacts through the Extended Data (ED) Collection feature. More details (opens new window)

• Optimized the YcFinServlet to resolve performance issues caused by the AppLog pattern logic.

• Added support for large log files on the AppLog page.

• Fixed various bugs in the AppLog module.

• Improved UI/UX on the FastThread Report page.

• Resolved ZCG parsing issues in GC log analysis.

• Configured /saml-acs.jsp as the default ACS (Assertion Consumer Service) URL for Single Sign-On (SSO).

• Minor bug fixes and improvements

yc-360 Script:

• Renamed the yc-agent script to yc-360 Script.

• Added a new argument httpClientTimeout to configure the maximum wait time for each HTTP request to the yCrash server.

• Enhanced the Extended Data (ED) capture capabilities. More details (opens new window)

• Minor bug fixes and improvements

3.0.9.3

yc-360 Script:

• Improved compatibility with Docker/Kubernetes environments, enabling successful capture of process, resource, and GC data.

• Resolved issue causing empty Process Reports due to incomplete capture of the top command output.

• Fixed missing Resource Reports by correcting the handling of vmstat command execution.

• Addressed GC log display failure by ensuring reliable execution of the ps command.

3.0.9

Server

• Log Masking Configuration: Introduced the -DmaskYCLogs system property to control masking of sensitive data such as IP addresses and S3/GCS bucket names. Masking is enabled only when the property is explicitly set to true (-DmaskYCLogs=true). If the property is unset or set to any other value than true, masking remains disabled by default.

• HeapHero Improvements:

  • Resolved CPU spike issues.
  • Enhanced the "Problem Detected" section in the heap analysis report.
  • Improved the "Potential Leak" page for better clarity.
  • Fixed minor bugs in the heap report rendering.

• Logging Enhancements: yCrash now logs at the ERROR level by default to the yc.log file.To enable detailed logs, set -DlogLevel=info in the startup script. Introduced a new property -DlogOutput=CONSOLE to redirect logs to the console (stdout) instead of a file.

• Storage Management: The auto-delete feature now enforces a 7-day default retention period for local data when remote storage is configured.

• Sample Configuration Files: Added sample files in the sample-conf directory for customizing thresholds, discarding incidents, configuring custom log4j settings, etc.

• Application Log Enhancements:

  • Improved pattern recognition logic and resolved related bugs.
  • Added server-side pagination to the applog table for faster loading.
  • Critical errors are now highlighted on the applog page for better visibility.

• (Alerts and Notifications: Integrated Microsoft Teams incoming webhooks into the workflow configuration for alert delivery.

yc-360 Script:

• Security Updates: Upgraded stdlib and x/net modules to address Go module vulnerabilities.

• Process Management: Resolved an issue that caused duplicate processes during M3 execution.

• Platform Support: Implemented automatic app log discovery for macOS systems.

• Codebase Improvements:

  • Refactored logic related to hd-sub and heap management for improved maintainability.
  • Added support for extended agent runtime duration.

3.0.8

Server

• Removed duplicate H2 DB JAR files that were causing compatibility issues in certain environments.

• Fixed an issue with discard settings; M3 will no longer log incidents related to discarded reasons.

• Improved clarity of problem statements in the heap report.

• Resolved bugs related to application log parsing.

3.0.7

Server

• Applog Page Optimization: Improved the loading performance of the Applog page, which previously experienced delays in data loading.

• Applog Repetition Pattern Enhancement: Enhanced the handling and analysis of repetition patterns in Applog.

• Heaphero Enhancements:

  • Added support for the 7zip file format.

  • Introduced an option to view threads during heap dump analysis.

  • Included a "Duplicate Classes" section in the heap report.

  • Enhanced the masking performance report in heap dump analysis when heap dump sanitization is enabled.

  • Optimized the performance of incoming and outgoing object references in the heap report.

• Improvements and Bug fixes

3.0.6.1

Server

• Resolved issues with table and index creation queries in the PostgreSQL database.

3.0.6

Server

• Application Health Check: Added a mechanism to verify whether the application is responding within the configured timeout. For setup details, refer to: Application Health Check (opens new window) Documentation.

• M3 Dashboard Fixes: Resolved bugs in the M3 Dashboard.

• AppLog Enhancements:

  • Detection of repeating statements.

  • Time gap analysis for better insights.

  • Enhanced error analysis.

  • Revamped UI.

  • Introduced an "Observation" section on the main app log report page.

• Pattern Recognition: Implemented basic pattern recognition in AppLog.

• Additional Log Support: Now supports GC logs, thread dumps, and heap dumps in 7Zip format.

• Heap Report Updates in HeapHero:

  • Added a "Threads" section to display all captured threads in the heap dump.

  • Introduced a "Duplicate Classes" section to identify classes loaded by multiple class loaders.

  • Added an option to sanitize heap dumps before analysis. More details: Sanitize Heap Dump (opens new window).

  • Enhanced sorting functionality in the largest objects and histogram tables.

  • Enabled viewing thread details directly from the largest objects table for objects of type java.lang.Thread.

  • Displaying JVM options in the Overview section of the heap report.

Agent

• Log Capture Enhancements: Refactored the capture mechanism for both application logs and system logs.

3.0.5

Server

• Resolved the SLF4J binding conflict.

Agent

• Security Update: Fixed a vulnerability in the stdlib go-module.

3.0.4

Server

1. Fixed yCrash Teams Issues:

   • Resolved an issue where the common team was not being created when the -DstrictAuthorization=false system property was configured, and the logged-in user was not part of any AD group.
   • Restored backward compatibility for existing users by ensuring that a new team with the licensee's name is created when remote storage is used.

2. Introduced an Appnames dropdown in the M3 Dashboard for easier application selection.

3. Memory usage is now displayed as a percentage in the Processes Report for Windows OS.

4. Added Boomi Charts and Reports for enhanced data visualization.

Agent

1. Security Update: Fixed a vulnerability in the Golang /x/net library.

2. Boomi Integration: The yCrash agent can now retrieve data from the Boomi server.

3.0.3

Server

1. Network Report Alerts:
   Alerts are now triggered when a host has over 20 total connections and any connection state exceeds 20% of the ESTABLISHED state.

2. M3 Dashboards:
   Various bugs in the M3 Dashboards have been resolved.

3. yCrash Teams Enhancements:
   Discover the latest improvements at yCrash Teams Documentation:

   a. A new user role, “super admin,” has been added to yCrash Teams.
   b. yCrash Teams now supports integration with AD/IdP groups.
   c. The -DstrictAuthorization system property has been introduced to manage user access.
   d. The -DsuperAdminGroupName property has been implemented to configure the super admin group name.
   e. You now have the option to restrict yCrash access to specific AD/IdP groups via the sso-authorized-groups.json file.
   f. Issues with the “my team” page in yCrash have been fixed.

4. Heap Report – Memory Leak Descriptions:
   The heap report now displays user-friendly descriptions for memory leaks.

5. Heap Report – Sorting Capabilities:
   Sorting functionality has been added to both the Largest Objects and Histogram tables in the heap report.

6. SAML Response Email Parsing:
   The logic for parsing the “email” attribute from SAML responses has been enhanced, and this attribute is now required.

3.0.2

Server

1. Postgres Database Support for M3 Metrics and Incident Storage: You can now use the Postgres Database to store M3 metrics and incident information. To configure Postgres on your server, refer to the setup guide: Postgres Database Setup Documentation.

2. Enhanced "My Teams" Feature with SQL Server and Postgres SQL Support: The "My Teams" feature has been upgraded to support Microsoft SQL Server and Postgres SQL, improving collaboration and team management. These enhancements enable smoother coordination, seamless information sharing, and faster issue resolution.

3. New BlockedThreadsSize Threshold Property: We’ve introduced a new threshold property, blockedThreadsSize, to monitor the maximum number of blocked threads. If the count exceeds this limit, an alert will appear in the Root Cause Analysis (RCA) report, helping you quickly identify and resolve performance bottlenecks caused by thread blocking. For details, see the documentation: Threshold Settings Documentation.

Agent

1. Enhanced Namespace and Podname Reporting in M3 Mode: The yCrash agent running in M3 mode now supports displaying the Namespace and Podname of target applications within a Kubernetes cluster. Note: To enable this feature, add kubernetes: true to the yCrash agent configuration YAML file.

2. Go Package Update: Go packages have been upgraded to version 1.23, addressing known security vulnerabilities.

3. Docker Base Image Update: The Docker base image has been updated to include the latest security vulnerability patches.

3.0.1

Server

• Upgraded the guava library to 33.3.0 from 19.0 to fix vulnerabilities.

3.0

Server

1. HeapHero Refinements:

   • Improved functionalities for better user experience.

2. Team Management Feature:

   • Introduced a new team management feature, available only with SSO integration.

3. Native Memory Tracking (NMT) Support

   • Added support for NMT reports, providing detailed insights into JVM native memory usage.
   • Features interactive graphs and alerts to facilitate efficient memory management and troubleshooting. Learn more Understanding Native Memory Tracking in Java (opens new window)

4. Settings Reader:

   • Added a comprehensive Settings Reader feature to view runtime configuration settings, formatted outputs, and JVM arguments via the /settings page (e.g., http://localhost:8080/settings (opens new window)).
   • Learn more Settings Reader Documentation

5. Enhanced Notifications UI:

   • Improved user interface for notifications.

6. My Teams Feature:

   • Introduced the My Teams feature to enhance collaboration and streamline team management, enabling easier coordination, information sharing, and faster issue resolution.

7. Heap Substitute Upload Option

   • Added the ability to upload a heap substitute within the individual Upload section for more flexibility.

8. Boomi Integration:

   • yCrash now supports integration with the Boomi enterprise server via the YC Agent.
   • yCrash Agent captures 360° data, including Boomi information, which can be accessed via the Boomi dashboard on the yCrash server. Learn more Boomi Integration Documentation

9. Discard Statements in RCA Reports:

   • Enhanced RCA reports to display discard statements for better issue analysis.

10. Custom Cooling Period:

    • Introduced a Custom Cooling Period to avoid repetitive incident reports for the same performance issue on a single server within a 24-hour window, while still logging issues from different servers. Learn more Custom Cooling Period Documentation

11. Observation Period:

    • The observationPeriod setting prevents false positives caused by temporary CPU or memory spikes by monitoring metrics over a set duration (e.g., 2 minutes).
    • yCrash captures 360° artifacts only if the spike persists, ensuring accurate performance monitoring. Learn more Observation Period Documentation

12. Enhanced the Dashboard widgets

13. Enchanced UI/UX and minor bug fixes.

Agent

1. Improved M3 Application Log Discovery:

   • Resolved performance issues in the M3 application log discovery process.

2. Updated Go Libraries:

   • Upgraded Go libraries to version v1.22.8 for enhanced compatibility and performance.

2.20.6.5

Server

• Added support for passing the encrypted SSL keystore password via the launch script. For more details, see the documentation

• Resolved an issue with the start and stop scripts where the process termination logic was delayed, improving graceful shutdown performance.

• Improve startup script for yCrash server

  1. Enhanced error handling for directory navigation and script execution.
  2. Replaced netstat with lsof for modern and reliable port checks.
  3. Added validation for extracted port numbers from the launch script.
  4. Improved readability and structured output messages for better clarity.
  5. Ensured robust process verification and log management after launch.
  6. Consolidated safety checks for compatibility across Linux and macOS.

2.20.6.4

Server

• Added support for configuring an encrypted javax.net.ssl.keyStorePassword in the launch script. To encrypt the password, please refer to the instructions provided at here

2.20.6.3

Server

Enhancements:

• Enhanced the security of keystore password management by introducing new methods for storage and retrieval. The following options are now supported:

  • Environment Variable: Allows setting the keystore password as an environment variable, reducing the need to hardcode sensitive information.
  • Hidden File: Enables storing the keystore password in a hidden file for added security.

These enhancements offer more flexible and secure ways to manage sensitive credentials within your environment.

Bug Fixes:

• Resolved an issue that prevented resources from being restricted against unauthorized access.

2.20.6.2

Server

• Implemented an option to bypass SSO authentication for yCrash REST APIs such as GCeasy, fastThread, and HeapHero APIs.

2.20.6.1

Server

• Resolved memory leak issue related to the shortest path calculation to the accumulation point.

2.20.6

Server

• Introducing yc-codec.jar encrypt configuration files and system properties. This jar is bundled in the yCrash build. After unzipping the yCrash build you can locate this jar file in tools/codec directory. To learn more about this jar please visit here

• Implemented a free memory validation logic for heap dump analysis. The tool will return an error if there is no sufficient memory to process the heap dump file.

• Bug Fix:

  • Handling encoding for the HeapHero response
  • Fixed memory leak detail page loading issue when common/shortest path to accumulation point has a lot of data

• Restructured HeapHero REST API JSON response. Renamed leakSuspect object to potentialMemoryLeak, and dominatorTree to largestObjects.

• Support to invoke GCeasy, fastThread, and HeapHero REST API with SSO authentication. To learn more please visit here

2.20.5

Server

• Thread Dumps REST API: The response now includes CPU-consuming threads.

• GC Event Handling: Previously, if the first GC event was several thousand lines into the log file, the tool would skip the entire file. This has been resolved by filtering out irrelevant data, ensuring that critical GC events are captured, even if they appear later in the log.

• Third-Party Library Detection: Minor fixes have been implemented to improve the accuracy of identifying and reporting third-party libraries responsible for slowdowns in thread dump analysis.

• Heap Dump Analysis: Performance improvements have been made to heap dump analysis.

• Heap Report UI Changes:
  a. The "Potential Memory" section has been moved to a separate page, accessible via hyperlink from the problem report.
  b. A progress bar has been added to the UI while problems are being identified.
  c. Hovering over segments in the sunshine chart now displays actual data.
  d. "Home" and "Back" icons have been added to the inner circle of the sunshine chart.

• HeapHero: Bug fixes implemented.

• GCeasy and fastThread: Bug fixes implemented.

2.20.4.1

Server

• Corrected the issue where h2-connector.xml was not being located properly when uploadDir was configured to a different directory.

2.20.4

Server

• Added a system property to enable monitoring of only Macro metrics in M3 mode. The "Monitor Only Macro Metrics" feature in yCrash helps prioritize critical performance metrics such as CPU and memory usage, reducing noise from minor micro metric fluctuations. Learn more about configuring the GC Monitoring in M3 and Thread Dumps Monitoring in M3

• Enhanced the largest objects chart view in the heap report with refinements and bug fixes.

• Introduced a new system property, “showGCReadyObjects,” allowing you to view objects that are no longer referenced by the application. This feature helps identify potential memory leaks and optimize memory usage. Learn more at Application Arguments.

• Improved the heap report UI for a better user experience.

• Refined the logic for detecting OutOfMemoryError:Metaspace, ensuring accurate forecasting of metaspace-related memory leaks through GC log analysis.

• Implemented minor bug fixes for improved stability.

• Updated the Helm chart for Kubernetes and OpenShift deployments to enhance compatibility and functionality.

2.20.3

Server

• The "Restricting Resources Access" feature has been introduced in yCrash to limit access to specific resources. Only administrators with a valid password can view these restricted resources. For more information, please refer to the following document.

• An interactive graph has been implemented to visualize objects in memory within the heap report, enhancing user experience and providing better insights into memory usage. To learn more about the interactive graph, read our blog here (opens new window).

2.20.2

Server

• Integrated MS SQL Server to store incident information in a database table. Please follow the steps provided here to configure the MS SQL Server database.

2.20.1

Server

Enhanced calendar dashboard performance by caching data into the H2 database. The H2 DB server can now be started in two ways:

Server Mode:

• Run the H2 server separately to allow communication with multiple yCrash servers behind a load balancer.
• Follow the instructions in this document to run the server in this mode.

Embedded Mode:

• No additional configuration is needed on your part; yCrash will automatically handle this mode.

2.20

Server

• Integrated a new mathematical model to detect instances of memory over-allocation in applications. Upon detection, the tool will recommend specific percentage reductions to optimize memory usage and enhance performance.

• Added support for parsing thread dumps generated by jcmd 22 in text format, enabling more comprehensive analysis and troubleshooting.

• Extended support to include parsing thread dumps generated by jcmd 22 in JSON format, improving the tool's versatility and usability.

• Introduced a new system property -DlogLevel to control log verbosity. For instance, setting -DlogLevel=error will log only error messages.

• Added a new threshold parameter, socketReadThreadsSize to define the maximum allowable number of socket threads. If this threshold is exceeded, an error message will be generated, preventing excessive resource consumption and potential performance issues due to too many open socket connections.

• Enhanced dashboard calendar loading performance by caching data into the database.

• Added flexibility to create custom log4j configuration files. Specify the file path as a system property using: -Dlog4j.configurationFile=<log4j.xml filepath>.

• Implemented various bug fixes in the REST API.

Agent

• Enhanced the appLogs logic in m3 mode to prevent uploading the entire file upon first encounter.

• Updated m3 mode to support the storagePath argument.

2.19.8.4

Server

• Upgraded the below libraries to fix vulnerabilities:
  1. bcprov (bouncycastle)
  2. webapp-runner

2.19.8.3

Server

• Resolved vulnerabilities caused by the commons-compress library

2.19.8.2

Server

• Addressed memory leak problem associated with thread creation and failure to terminate in the calendar builder module.

2.19.8.1

Server

• Enhanced stability of the implementation for the -DlogOutput system property.

• Updated the name of the "Leak Suspect" section in the heap report to "Potential Memory Leak".

• Revised display of potential memory leaks to only appear if the object is occupying more than 200 MB of memory due to leakage.

• Resolved the issue with SSO redirection.

• Optimized performance of the dashboard calendar page.

2.19.8

Server

• Improved error handling for HTTP status codes 401, 404, 500, 501, and more.

• Enhanced GC log parsing to support multiple generations in the ZGC algorithm.

• Addressed minor bugs in the parsing logic for Java 9 G1 GC logs.

• Made slight improvements to label readability in GC log graphs.

• Resolved issues with upload and report counts in GCeasy/FastThread dashboard RestAPI.

• Updated documentation for the Agent Setup page and Trial Instruction page.

• Included all APIs and JSP pages under SOS.

• Strengthened security measures for uploaded file formats.

• Added a new system property -DlogOutput=CONSOLE to configure console output.

2.19.7

Server

• Enhanced stability of the HeapHero UI.

• In the HeapHero cloud, troubleshooting reports are now loaded by default, while performance reports are hidden.

• Introduced a new parameter "version" in the /debug API to exclusively return the version number. For example: /debug?version.

• Refactored and optimized the auto-delete incident feature for improved performance.

• Addressed bugs in the auto-delete incident feature to ensure smoother functionality.

2.19.6.1

Server

• Enhanced the stability of the auto-delete incident feature.

2.19.6

Server

• Resolved the usage-stats issue to include the username for the debug API.

• Enhanced heap dump analysis to reduce memory consumption, particularly in the following sections Largest objects and Leak suspects

• Addressed the cooling period issue in M3 data capture.

• Implemented the deletion of incidents when M3 detects problems that do not exist in the 360 data capture.

Agent

• Enhanced timezone detection within M3 data capture for improved accuracy.

2.19.5.1

Server

• Resolved the Bundle Upload and Individual Upload issues.

2.19.5

Server

• Implemented logic to detect errors reported in the application log occurring at specific frequencies. The 'count' and 'timeWindow' parameters have been added to the log-error-settings.json file.

• Developed functionality to identify recursively looping threads causing CPU spikes.

• Improved the stability of heap dump analysis and the heap report user interface.

• Implemented scheduling of auto-delete CRON job upon server startup.

• Users logged in using Saml authentication can now view their username in the profile avatar.

• Resolved instances where M3 detected problems not present in the 360 data capture. Incidents will now be removed from the dashboard.

2.19.4

Server

• Implemented functionality to parse NodeJS Garbage Collection logs.

• Optimized algorithms to minimize false incident detections during Micro-Metrics Monitoring (m3).

• Improved CPU spike detection algorithm for better accuracy.

• Enhanced stability of Metaspace Memory Leak detection logic.

• Increased logging to monitor user activities in the "usage-stats" file when using SSO.

• Enhanced UI stability for the heap report page.

• Expanded auto-delete feature to support additional time formats such as days, months, and years. For more information about deleting incidents, refer to the following link: Auto-Delete Incidents.

Agent

• Corrected the issue with transmitting heap dump files to the yCrash server when utilizing IBM/OpenJ9 JVM.

• Implemented support for the "%seq" parameter in the GC log file name.

2.19.3.3

Server

• Fixed notification issue when the yCrash agent transmits data to the yCrash server.

2.19.3.1

Server

• Resolved the issue related to "remote location" in the HeapHero REST API.

• Addressed the redundancy of waiting thread error/warning messages in the RCA report.

• Fixed an issue where hyperlinks were not displaying in the RCA report for thread dump uploads during web uploads.

2.19.3

Server

• Implemented logic to detect memory leaks in the metaspace memory region.
• Excluded the hot.jar script from the codebase.
• Discontinued the highlighting of log statements containing error/exception keywords, even when the log level was DEBUG or WARN, to prevent false reporting.
• Resolved the issue of duplicate error reporting in thread dump analysis.
• Omitted the reporting of Max GC time in the GC log analysis report when parsing jstat log output, as the jstat log output does not contain Max GC pause time metrics.
• Addressed various bug fixes in yCrash reports.
• When uploading logs from cooltools web applications or individual uploads, filenames will now be displayed as tags.
• Log Error Detection Settings: The application log file records various events, errors, and information during application execution. The yCrash agent monitors the application log for critical errors. When critical errors are reported in the application log, the yCrash agent captures 360° troubleshooting artifacts from the application stack (GC log, thread dump, top, netstat, vmstat, dmesg...) and sends them to the yCrash server for analysis. By default, Stackoverflowerror and OutofMemoryError have been added to global settings in log error settings to capture 360-degree data whenever an error related to Stackoverflowerror or OutofMemoryError occurs.
• Resolved the issue of redundant waiting thread error/warning messages in the RCA report.

2.19.2.4

Agent

• Security vulnerabilities have been addressed by updating the Go base to version 1.21 in this release.

• All module libraries have been upgraded and patched to ensure security and stability.

2.19.2.3

Server

• Addressed vulnerability issues in spring-core, spring-web, spring-test and tukaani-xz dependencies.

2.19.2.2

Agent

• Fixed vulnerability issue in https://gopkg.in/yaml.v3 (opens new window) package of yCrash agent.

2.19.2.1

Server

• Upgraded the following libraries to resolve vulnerabilities in the yCrash application.

  httpcore-4.4.15.jar
  jackson-dataformat-cbor-2.14.2.jar

2.19.2

Server

• Upgraded the following libraries to resolve vulnerabilities in the yCrash application.

  zip4j-2.11.3.jar
  commons-io-2.5.jar
  org.apache.httpcomponents_httpclient
  commons-codec
  tika-parsers-1.28.5.jar
  webapp-runner.jar
  guava-32.1.3-android.jar

2.19.1

Server

• Upgraded the following libraries to resolve vulnerabilities in the yCrash application.

  spring-web-5.3.31
  aws-java-sdk-s3-1.12.261
  aws-java-sdk-1.12.261
  aws-java-sdk-core-1.12.261
  aws-java-sdk-ses-1.12.261
  commons-fileupload-1.5
  google-oauth-client-1.33.3
  gson-2.8.9
  guava-24.1.1-android
  httpclient-4.5.14
  jackson-annotation-2.16.1
  jackson-core-2.16.1
  jackson-databind-2.16.1
  json-20231013
  mysql-connector-8.0.33
  newrelic-6.5.0
  spring-web-5.3.31
  spring-core-5.3.31
  spring-test-5.3.31
  tukanni-xz-1.9
  xmlsec-4.0.0
  tika-parser-1.26

• Optimized memory consumption in HeapHero.

2.19

Server

• Added a new property for ThreadCpuPercentage (also known as threadConsistentCpuPercentage) in both cloud and EE.

• Included a Mac-specific tab in the ycrash trial instructions page and ycrash agent setup page.

• Modified the "What's in Memory" section to present information in a more human-readable format.

• Integrated Slack webhook notifications, allowing real-time incident alerts and data from yCrash Server to be sent directly to your Slack channel.

• Added additional logs to troubleshoot issues related to making HTTP calls and checking server availability.

• Made the auto delete job triggering configurable by introducing a new property called "runJobAt" in the auto-delete-frequency.json file.

• Introduced the property "localTempRetentionPeriodInHours" for local storage auto delete.

• Addressed UI-related issues in the AppLogs report.

• Improved handling of multiple thread dump reports in yCrash.

• Strengthened error/problem detection by analyzing application logs. This enhancement helps detect OutMemoryError, StackOverflowError, and user-defined errors mentioned in the log-error-settings.json.

• Strengthened memory leak detection logic specifically for the Azul Zing GC algorithm.

• Fixed issues related to duplicate error reporting when threads are transitively blocked in the thread report.

• Resolved parsing issues related to garbage collection (GC) causes.

• Introduced the ability to specify the application version in ycrash EE.

• Addressed minor bugs in yCrash.

Agent

• Enhanced the logic for capturing application logs.

• Refactored the GC log file capture to accommodate various file patterns in the Windows environment.

• Updated M3 metrics capture logic for CPU and MEMORY in Windows.

2.18.5

Server

• Fixed OOM (OutOfMemoryError) issue when uploading Heap dump file.

2.18.4

Server

• Improved the stability of heap dump upload logic and resolved an issue with heap dump upload.

2.18.3

Server

• Stabilised heap dump upload functionality to improve the upload speed.

• Fixed heap dump file upload validation issues.

2.18.2

Server

• Integrated MS Team with "Alerts and Notifications".

• Made changes to accelerate heap dump upload speed.

• Simplified heap report page and fixed bugs.

2.18.1

Server

• Enhanced the Discard settings to support Regular Expression & made changes to the discard-settings.json structure to support the same.

• Thread dumps generated from WebMethods console has a special type of timestamp. Now we are parsing that format of timestamp and printing in the left panel menu item and report header

• Enhance yCrash Agent setup page and yCrash trial instructions page to make the setup instructions more granular.

• Introduced HeapHero API upload section and bug fixes

• UI bug fixes in the App log report.

• Bug fix in creating the first incident when M3 launched.

• Fix minor bug fixes in hs_err_pid processing.

Agent

• Resolution for capturing GC logs with placeholders %p and %t in the file path has been implemented.

• The default setting for appLog line counts has been adjusted to retain the last 2000 lines.

• Timezone capture for m3-fin has been integrated.

• Application logs are auto-discovered. As part of this auto-discovery, GC logs were also discovered and parsed. This caused duplicate analysis of GC log. This is now fixed.

2.18

Server

• Introduce a new feature called My patterns (opens new window) in thread report. This feature will let the user add their custom pattern to matched in the thread dump. If they are matched, then it will be flagged as alert.

• Introduce new settings in app settings page (Parallel Parsing, Only Troubleshooting report, only Performance report)

• Time filtering GC log files was in beta version in the last release. Made several bug fixes and strengthened its implementation to handle complex cases.

• Major enhancements and bug fixes in the M3.
  a) Not Triggering incident during Cooling Period.
  b) At the very first time agent will create the first incident in the calendar.
  c) Transient Problems tripping m3
  d) Bug fixes in UI.
  e) Changed throughput chart to memory usage chart in M3 dashboard.
  

• Added capability to upload dump files from the individual report itself directly.

• Strengthen upload heap dump validation in HeapHero.

• Strengthen implementation of -DonlyTroubleshootingReport and -DonlyPerformanceReport properties in HeapHero.

• Renamed dominator tree section to largest objects.

• Changed the UI design of the heap report page. By default, it displays only the overview, leak suspect, and largest objects section in the troubleshooting report. The performance report will be collapsed by default.

• Added support for read heap dump files which are compressed in '.tar.gz' file format.

• Made several Bug fixed in HeapHero

• FastThread tool wasn’t processing certain types of hs_err_pid log. Now it has been fixed

• Added Potential solutions for hs_err_pid log.

• Enhancements in Divide and Conquer panel

Agent

• Adjusted M3 Frequency to align with the complete 360-degree artifact capture cycle.
  – Previously, the capture cycle wasn't aligned with the "m3Frequency" settings, resulting in captures occurring only after 5 minutes if the frequency was set for intervals shorter than 5 minutes. During other frequency intervals, it exclusively captured GC logs and TOP data, omitting a comprehensive artifact collection. This adjustment ensures a complete capture cycle within the specified m3Frequency settings.

• Agent versioning will now synchronize with the Server release cycle.

• Avoiding full GC in the application while capturing heap substitute data from the application.

2.17.8

Server

• Handling the scenario where multiple files are present in a thread dump and each file contains Top data first followed by Thread dump.This scenario was working properly when uploaded directly from the fastThread tool. However it was failing to work when uploaded from 'Individual Uploads'. Now fix it as well.
• Upgraded the spring-core library to new version 5.3.30

2.17.7

Server

• Resolved the issue of multiple thread dumps in the left navbar.
• Fixed the problem of leak suspect loading with remote storage.

2.17.6

Server

• Upgraded the webapp-runner library to fix the vulnerability issue reported by the customer.

2.17.5

Server

• Improved compatibility of "start-yc-server.sh" and "stop-yc-server.sh" scripts with various operating systems.

Agent

• Enhanced support for capturing GC log files with the %t naming convention.
• Deprecated the use of the '-cmd' and '-urlParams' options.

2.17.4

Server

• The issue of TD Timestamps not printing correctly in wrapper.log has been resolved.
• The issue of CPU stats not being reported has been fixed.
• The issue of AIX - jstack generated thread dump not being parsed properly has been resolved.

2.17.3

Server

• Enhance the application log parsing to print the error message along with the exception name in the Error graphs
• Fixed the Leak Suspect section load issue when agent transmits and when the RestAPI is invoked.
• The left side navigation bar in thread report and tables in comparative summary report now display timestamps.
• Fix a few minor app log parsing errors
• Simplify Agent Setup documentation.

Agent

• Fixed gcPath discovery issue on Windows agent.

2.17.2

Server

• The errors in the yc.log file have been reduced/silenced.

2.17.1

Server

• The issue with the licensing analysis count has been resolved.

2.17

Server

• We have fine-tuned our patterns and machine learning algorithms to categorize CPU-consuming threads with appropriate priority.
• The ‘Divided & Conquer Panel’ feature has come out of beta and is now live in this release. This feature categorizes identical stack traces, facilitates you to move threads to the appropriate tabs, and ultimately identify problematic threads with ease. For more details, refer to https://blog.fastthread.io/2023/09/18/divide-conquer-problematic-threads
• 'My Patterns' is a new feature that empowers you to create custom patterns that define specific thread stack trace structures or sequences. These patterns are the key to focusing your analysis on exactly what matters to you. Think of them as your precision lenses. For more details, refer to https://blog.fastthread.io/2023/09/18/harnessing-precision-in-thread-dump-analysis-with-my-patterns
• We have introduced ‘Filter By Time’ in GCeasy, which is now in beta stage. This powerful tool allows you to precisely analyze your Java GC logs for specific time periods. With this tool, you have the ability to tailor your reports and metrics, focusing on the exact moments that matter most for your application’s performance.
• We have introduced an ‘API’ tab in both the GCeasy and fastThread upload widgets to highlight the API key and end-point URL.
• We have fixed Notification UI bugs.
• We now show CPU threads that are above 40% only. If the CPU percentage is between 40-80%, we make it a warning. Above 80%, it is a danger.
• We calculate the average of particular CPU threads and show its average in the Summary Report.
• We have added the ability to tag Alerts and Notifications in this release

Agent

• The current output of the yCrash agent is directed to the standard output. However, it has been modified to direct the output to the standard error. By doing so, the yCrash agent output can be redirected to a log file using the “2>” option.

2.16.3

Server

• To enable the download heap dump option on the heap report, you can add a system property "-DdownloadDumps=true" to the launch script.
• The performance of the leak suspect section in the heap report has been improved.
• The issue of loading the reference page of an old heap report has been fixed.

2.16.2

Server

• Heaphero: We fixed an issue where the child nodes are not showing up when we clicked on a parent node in the Outgoing Reference page of Dominator Tree.

2.16.1

Server

• Introducing Object Query Language (OQL) capability for advanced heap dump analysis.
• Added the option to view "Common path to the accumulation point" in leak suspect analysis for more comprehensive heap dump insights.
• Enhanced the validation logic for heap dump file uploads to ensure data accuracy.
• Implemented the ability to choose between Troubleshooting Report and Performance Report viewing modes.
• Improved error handling by triggering 360-degree data capturing in critical error situations.
• Metadata like application name, host IP, and tags can now be passed in GCeasy, FastThread, and HeapHero APIs, with this information displayed in the yCrash calendar dashboard.
• Introduced heap substitute file uploads in individual uploads for enhanced analysis.
• Enhanced notification capabilities: Get alerted about new incidents via WebHook events or Google Chat.
• Added a new 'Fatal' severity level in thread dump analysis API for identifying critical errors affecting application availability.
• Expanded GC log analysis API with 'fatals' and 'warnings' elements to identify and address memory problems and performance concerns.
• Improved RCA report by removing duplicate elements in blocked threads.
• Added an animation effect in the "Eliminate and conquer" panel for a more engaging user experience.
• New application launch scripts (start-yc-server.sh and stop-yc-server.sh) for Linux and macOS environments.
• Shell script files for Linux and macOS are now executable by default for convenience.

Agent

• Added support for compressed log files in the appLogs section with .zip and .gz extensions.
• Resolved a critical issue related to processTokens support in the Windows environment.
• Enabled auto discovery of application logs in the Linux environment.

2.15.1

Server

• Fixed the large attribute data load issue in the HeapHero actual data table.

2.15

Server

• Introduced a new feature eliminate and conquer panel in thread dump analysis report.
• Streamlined the parsing of the hs_err_pid file in the fastThread application
• Upgraded UI of yCrash Dashboard
• Introduced the "-DhdParallelParsing=true" system property to configure yCrash to parse heap dumps in parallel.
• Added a new HeapHero REST API request parameter "performanceReport=true" to retrieve performance report data in REST API responses.
• Implemented a mini progress bar on the heap report page, providing background progress visibility while analyzing the heap dump.
• Added a mini progress bar on the RCA report page to indicate heap dump file parsing progress.
• Added support for .dump, .dmp, and .mdmp file extensions for heap dumps.
• Resolved an issue with decoding Unicode characters in the troubleshooting report.
• Fixed a memory usage problem caused by not closing the snapshot object.
• Addressed other minor bugs.
• Added more recommendations to GC log analysis
• Added videos explaining the bundle upload, individual upload, and remote capture features for better understanding.
• Bug fixes in multiple application logs parsing.
• New UI improvements in M3(Micro metrics monitoring) pages.

Agent

• Fixed an issue to support multiple log files in appLogs with the same filename.
• Added support for GC logging with IBM-J9/OpenJ9 JVMs.
• Improved filename parsing with appLogs in onlyCapture mode.
• Resolved a bug related to the vmstat capture command.

2.14.1

Server

• Fixed Unicode character escaping issue in the troubleshooting report table.

2.14

Agent

• Improved automatic retrieval of GClog flag from Oracle Hotspot JDK 11 JVMs.
• Enhanced appLogs functionality for better support.

Server

• Rearranged tabs in the heap report for improved usability.
• Added progress bar for heap dump file uploads.
• Bug fixes and enhancements for appLogs.
• Usability defects fixed in yCrash.
• Enhanced CPU & Memory section in fastThread.

2.13.1

Server

• Fixed thread stacktrace page loading issue in heap report.

2.13

Agent

• Added support for vmstat parsing in Windows environment.

Server

• LogSense (Beta) release. Added capability to parse unstructured application log file contents into structured table view, list view format.
• Eliminate redundant CPU spike thread error messages in RCA report.
• Only if more than 5 threads are BLOCKED in a lock, then raising it as an alarm. This helps to silence false positive errors.
• Added support to parse the latest version of hs_err_pid log files.
• Started to report 'Process report' for windows environment.
• HeapHero: Implemented user-friendly error messages and recommendations.
• Auto Delete enhancements to delete logs, M3 metrics and other files.

2.12

Agent

• Supporting %P in gc log path in JDK 8 and above versions.
• Collecting cpuCount to avoid false alarms.

Server

• Based on the cpuCount of servers it will show alerts in the Processes report.
• Supporting MacOS Top.
• Bug fixes in Windows Processes report.
• Bug fixes in the kernel report.
• Export as PDF option added in Heap Report
• IBM heap dump support for troubleshooting report
• Download heap dump option in the Heap Report
• Memory usage was spiking up while analysing the heap dump. Now executing parsers in sequential order instead of running them parallelly.
• Compressed IBM heap dumps were failing to show the Heap Report
• Compressed file path in remote location feature was not displaying the Heap Report
• Fixed "See Thread Stacktrace" option the leak suspect of Heap Report
• Other minor bug fixes

2.11

Agent

• Dmesg logs support in multiple platforms Windows,Linux and MAC
• Minor bug fixes in Agent.

Server

• Fixed GC G1 parsing times issue.
• Implement Export PDF in Heap Report.
• Log message clean up.

2.10

Server

• Added "Effective Tips" link in CPU Consuming thread section.
• GC G1 parsing issue fixed.
• Threshold settings issue fixed.
• HD Substitute printing is changed in human readable format.
• Introduced a new "Training" menu in the header section and the "Blog" menu moved to under the Resources dropdown.

...