Release Notes 0.62 ------------------ yCrash Agent ------------- + Logging: You can configure yCrash agent's messages to be sent to a log file by passing '-logFilePath' argument. You can configure to roll over the log files if it goes beyond certain size by passing 'logFileMaxSize' argument. For more logging related arguments refer to https://docs.ycrash.io/ycrash-agent/all-agent-arguments.html + dmesg capturing: Starting from this release 'dmesg' data is captured from the device and transmitted to the yCrash server for analysis. + onlyCapture: New argument 'onlyCapture' has been added. Default value is 'false'. When this argument is passed with the value 'true', agent will capture all the artifacts from your application, compresses those artifacts, and creates a zip file () in the folder where it was run. But artifacts will not be transmitted to yCrash server for analysis. This flag will be useful if your application is deployed on the client location, and for security reasons, your client doesn't want to automatically transmit the artifacts to yCrash server running on your site. + URL endpoints printed in table format Once agent completes running, it will print all the API/dashboard report URLs in a table format in the console as shown below: +-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+ | Port | 8085 | | Status | success | | RunTime | 9.7573309s | | dashboardReportURL | http://192.168.1.70:8080/tier1app/yc-report.jsp?ou=buggycompany&de=192.168.1.70&app=yc&ts=2021-01-28T06-56-07 | | gcMetricsURL | http://192.168.1.70:8080/tier1app/yc-reader?ou=buggycompany&de=192.168.1.70&app=yc&ts=2021-01-28T06-56-07&dt=gc&apiKey=buggycompany@e094a34e-c3eb-4c9a-8254-f0dd107245cc | | threadMetricsURL | http://192.168.1.70:8080/tier1app/yc-reader?ou=buggycompany&de=192.168.1.70&app=yc&ts=2021-01-28T06-56-07&dt=td&apiKey=buggycompany@e094a34e-c3eb-4c9a-8254-f0dd107245cc | +-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+ + Reduced overall runtime: yCrash agent code has been optimized, such that it will complete capturing and transmitting all the data with in few seconds. + m3 mode heap dump capture: When yCrash agent is configured to run in m3 mode, if it detects problem it will capture the heap dump, even though '-hd' flag is specified as false. Now this issue has been fixed. Only if '-hd' flag specified with value 'true', heap dumps will be captured. yCrash Server -------------- + New tool to analyze top command output: 'top' is a popular Operating system command to report overall system's resource utilization. In the 'Cool Tools' section we have introduced a new tool 'Top command analyzer'. This tool will parses top command output and print the results in the browser. Tool identified various vulnerabilities in the output, and recommends solution to fix them as well. Public version of this tool can be found here: https://ycrash.io/yc-top-analyzer.jsp + Threshold feature yCrash application monitors several macro and micro metrics to detect the problems in the application. If these metrics values are breached beyond certain threshold then alerts are generated. Say for example default threshold value for maximum Garbage collection pause time is: ‘5 seconds’. If your application’s garbage collection pause time exceeds 5 seconds, then alerts will be generated. For certain applications (like trading application) ‘5 seconds’ might be a too high value, on the other hand it might be acceptable value for several other applications. Starting from this release, yCrash provides a mechanism to customize these thresholds values based on application’s need. TO learn more visit https://docs.ycrash.io/ycrash-server/administration/threshold-settings.html + Process report enhancement Process section in the incident report has been enhanced significantly. Following are the enhancements: a. New Graphs and visualizations are added b. Multiple error detecting rules has been added c. Resolutions/Recommendations sections are added to resolve the identified problems. d. Several help section documents has been added. + Compare Report enhancement Compare report dashboard has been enhanced. Now a date picker has been provided in the compare report page. User can select the incident reports for comparison by going through this date picker. It's lot easier than copying and pasting two different incident URLs manually. + '/debug' - refinements/excception handling '/debug' is the end point used by various customers to do health check of the tool. In this release this health check end point has been optimized for better response time. + Metaspace parsing logic fix In certain types of GC logs, overall metaspace size was wrongly reported in 'JVM Memory size' section. Now it has been fixed. + G1 GC log parsing For certain types of G1 GC events, gc event type weren't correctly classified. This issue has been fixed in this release. + Fix SAP JVM thread dump lock parsing logic Thread Lock information present in the SAP JVM generated thread dumps weren't parsed correctly. Now it has been fixed. + Network report refinements Graphs in network report has been changed for better readability. + Email Notification feature Email Notification has been enhanced. Now notifications can be sent certain email addresses only when incidents are reported on a particular application or hostname. For more details refer to: https://docs.ycrash.io/ycrash-server/administration/email-notification.html#step-1-smtp-host-configuration + dmesg capturig and reporting dmesg is a logging done by the Linux operating system to report any Operating system errors, warningg. Now this dmesg is captured by the agent and transmitted to the server for analysis. + Simplification for blocked threads BLOCKED threads are reported using transitive dependency graph. Graph were unreadable if multiple threads are BLOCKED. Now we are streamlined the graph - so that multiple BLOCKED threads are painted in concise manner. + Exception thread bug classification 'Exception Threads' section in the Thread report has been cleaned-up. Earlier certain threads were wrongly classified as exception threads. Now it has been fixed. + In certain GC reports, for certain data points, young gen and old gen sizes aren't painted properly in the graph. Now it has been fixed + Eclipse MAT errors When errors are reported in the Eclipse MAT report, those error count aren't updated in the Calendar dashboard. Now it has been fixed.