Crash
Download
Focus Mode
Font Size
RUM Pro provides comprehensive crash analysis capabilities, supporting the collection and reporting of various crash exceptions on both Android and iOS platforms to help users better address application quality issues. This article details the features and usage of each module within the crash analysis functionality of RUM Pro.
Crash Overview
On the Crash Overview page, RUM Pro provides comprehensive analysis of crash rates and crash trends, primarily supporting the following analytical capabilities.
Metrics Overview
Supports the display of today's Device crash rate, Crash rate, Occurrences, Number of affected devices, and supports day-over-day comparison relative to the previous day.
Device crash rate: Device Crash Rate = Affected Device Count / Connected Device Count.
Crash rate: Crash Occurrence Rate = Number of exception cases meeting filter conditions / Number of launches meeting filter conditions. (Currently, the statistics for launch counts only support two filter conditions: time range and App version.) Data is only available for viewing from July 29, 2023 onwards.
Occurrences: The number of occurrences of exception cases that meet the filter conditions.
Number of affected devices: The number of devices associated with exception cases meeting the filter conditions, with duplicates removed based on device ID during exception reporting.
Crash Trend
App Version: Contains all reported App versions. The dropdown allows selection of all versions or a single App version for filtering and viewing.
Time Granularity: Supports aggregation by 5-minute intervals, 1-hour intervals, and daily intervals.
Display: Trend charts can be viewed with coordinate dimensions set to percentage, per mille, or per ten thousand.
Statistical Dimension: Includes two dimensions: 5-minute and cumulative. The 5-minute dimension trend chart aggregates data every 5 minutes based on filter conditions, providing more granular time points. The cumulative dimension trend chart aggregates data every 55 minutes based on filter conditions, offering a more intuitive view of crash trends.
Supports multi-dimensional display of crash trends. At Time Granularity, select aggregation by 5 minutes, 1 hour, or by day. Simultaneously, in the right Tab bar, you can switch to display Device crash rate, average crash rate per devices, Crash rate, Occurrences, Launchs, Number of affected devices, Number of networked devices.
Supports the version comparison feature. Click App Version to select different application versions for comparison in the trend display.
Supports crash analysis for the Top Issues within the trend chart time range. In the trend chart, clicking on different data ranges in the curve displays the Top 3 issues and their impact scope for that range below.
Crash Rankings
Supports drill-down statistics on crash rates by app version, system version, and device model. Click the dropdown in the upper-right corner to switch between Number of affected devices and Number of networked devices.
Supports statistical display of the Top 20 crash issues on specified dates, including their impact scope and fluctuations.
Crash Problem List
Problem Query
In the crash problem list, RUM Pro supports comprehensive individual case query capabilities, primarily as follows:
Supports drill-down search through multi-dimensional fields such as time range, App version, user ID, device ID, foreground/background state, and Issue type.
Supports advanced search features for error stack traces, including tokenized and non-tokenized search, with field intersections, unions, or fuzzy matching capabilities.
Supports search for custom issue types such as processing status, assignee, Tag, and tag colors.
Supports search for business-reported Custom Fields.
Supports searching by attachment name within Problem case details, allowing filtering of cases with corresponding attachments reported.
Trend analysis
Problem List
The problem list supports problem management features such as feature overview, impact scope assessment, trend analysis, assignee tracking, and processing status management.
Statistical distribution
Supports drill-down statistics on crash rates by app version, system version, and device model. Click the dropdown in the upper-right corner to switch between Occurrences, Number of affected device, or Affected Users.
Crash Issue Details
Detailed case infomation
Error Details: Includes information such as Occurrence Time, User ID, system version, Model, Bundle ID, build number, Process launch ID, SDK version, Usage Duration, Runtime Architecture, Scene, Reporting Time, Device ID, Vendor, foreground/background, APP Version, Launch ID, Message ID, faulty process#thread, ROM, CPU architecture, and Channel. This provides a more intuitive display of detailed information about this exception report.
Recently reported: Displays the latest reported issue information, including APP Version, reporting time, Device ID, system version, and device model.
Report device: Sorted by reported device IDs, enabling better search for specific error cases based on device ID. Includes APP Version, reporting time, Device ID, system version, and device model.
Report user: Sorted by reported user IDs, enabling better search for specific error cases based on user ID. Includes APP Version, reporting time, User ID, system version, and device model.
error stack
The stack trace provides detailed information about the current crash issue and serves as the most direct information for analyzing and resolving crash problems.
In Java crash stack traces, the error stack displays the error message and the current call stack, which typically consists of file name + symbol + offset. If the Java code was obfuscated during compilation, you can upload the obfuscated mapping file to translate the obfuscated symbols. If the mapping file has been uploaded, use the button on the right to switch between displaying the Original stack or Restored stack, with the restored stack displayed by default.
If the current thread stack cannot pinpoint the issue, you can click Unfold thread stack to view other thread stacks. Other thread stacks also support symbol deobfuscation.
In the error stack, if the corresponding symbol table for the stack has not been uploaded, the page will indicate that the stack is untranslated. You can click Upload to navigate to the symbol table upload page. After uploading the symbol table, click manual restoration in the upper-right corner of the page. Wait a few minutes and refresh the page; the stack will then be deobfuscated. For the operation procedure, see Symbol Tables.
on-site data
Crash snapshot data captures the on-site information collected at the moment of a crash, primarily including the following content.
Basic Information: Includes Network APN, memory status, storage status, JVM status, etc. The description of JVM status is as follows:
Maximum JVM memory: Indicates the maximum heap memory size that can be used by the JVM.JVM allocated memory: Indicates the size of memory currently allocated to the JVM. Typically, the allocated memory is less than the maximum memory. When a Java program requires more memory, the JVM automatically expands the heap memory until it reaches the maximum limit. If the JVM cannot allocate additional memory, it may throw an OutOfMemoryError exception.JavaHeap: Indicates the portion of allocated memory used to store Java object instances.PSS: Indicates the size of the process's shared memory.VSS: Indicates the size of the process's virtual memory.custom dimension: Custom dimensions are user-defined data with constrained keys. Keys are defined by the SDK, and applications can set values with different meanings based on actual needs. Values currently support three types: Number, String, and String Array, with each type supporting ten keys.
Values of the Number type correspond to dimension keys n1 ~ n10.
Values of the String type correspond to dimension keys s1 ~ s10.
Values of the String Array type correspond to dimension keys a1 ~ a10.
Custom Fields: Display custom fields uploaded by users via the API, can refer to custom data.
Page tracking: Displays the lifecycle tracking of pages prior to a crash, allowing analysis of user page interactions before the crash occurred.
Log
The log section displays the log stream before and after the crash occurrence time, supporting search by log level and keywords.
FD info
FD info displays the FD usage status at the time of crash.
Includes the current number of used FDs, unused FDs, and the maximum number of FDs available per process. FDs are categorized by type, quantity, and proportion. The FD details display the specific paths of opened FD resources, with the number of FDs pointing to each path indicated at the end. FD information can assist in troubleshooting by identifying FD types that are excessively opened during crashes, enabling targeted optimization for these FD types.
Process Information
Note:
Native crashes report this information, while Java crashes do not report it.
Process information displays details related to the process at the time of crash, including Maps information, Meminfo information, process status, and thread status. The raw content of process information can be viewed in the attachment state_file.zip.
Maps information represents the memory image of a running user process, allowing visualization of the current process space mapping. Specific memory segments can be filtered by address, segment name, and access permissions. Memory segment filtering supports regex matching, while access permissions require exact matches (e.g., rwxp denotes a private memory segment with readable, writable, and executable permissions).
Meminfo information indicates the memory usage of user processes, showing how the application's current memory is primarily allocated. The coordinate descriptions in Meminfo information are as follows:
Coordinates | Field | Description |
X-axis | Native Heap | Local heap usage. |
| Dalvik Heap | Dalvik heap usage. |
| Dalvik Other | Memory usage outside the Dalvik heap. |
| Stack | Process stack usage. |
| Ashmem | Anonymous shared memory usage. |
| Other dev | Other device memory usage. |
| .so mmap | Shared libraries' memory usage. |
| .jar mmap | JAR file's memory usage. |
| .apk mmap | APK file's memory usage. |
| .ttf mmap | Font file's memory usage. |
| .dex mmap | DEX file's memory usage. |
| .art mmap | ART file's memory usage. |
| .oat mmap | OAT file's memory usage. |
| Other mmap | Other memory-mapped files' usage. |
| GL mtrack | OpenGL ES memory usage. |
| Unknown | Unknown memory usage. |
Y-axis | Pss Total | The total physical memory used by the process, including shared libraries, anonymous shared memory, heap and stack, etc. |
| Private Dirty | Private Dirty memory size used by the process, i.e., the memory size exclusively used by the process, including heap, stack, anonymous shared memory, etc. |
| Private Clean | Private Clean memory size used by the process, i.e., memory which has been cleared and contains no data. |
| SwapPss Dirty | Swap dirty page size of the process. |
Process Status and Thread Status indicate the state metrics of the currently running process and its individual threads, respectively. In Thread Status, you can click the arrow next to the thread name to expand the status of a single thread. When hovering over a metric, the description of the corresponding state indicator will be displayed.
Symbo Table
In the message details of a crash instance, you can also upload symbol mapping files. Unlike the symbol mapping file upload in Settings, this section only displays the required symbol mapping files for unresolved Java stacks and Native stacks in the current instance's error stack.
Java symbol table files are matched based on version number + build number, while so symbol table files are matched based on UUID. Click Upload Pending, click to upload to directly upload the corresponding symbol table on this page. After uploading, click Manual Deobfuscation to perform the deobfuscation. Wait a few minutes, and the current error stack will be deobfuscated. For symbol table related information, refer to the symbol table.
Attachment
RUM Pro supports package download for all attachments. User-customized reported data files can be found in the following attachments:
The content set via the callback API
getCrashExtraData is stored in userExtraByteData.The content set via the callback API
getCrashExtraMessage is stored in the user_datas.log file.The content set via the custom field
putUserData is stored in valueMapOthers.txt.The large files set via the custom file
setAdditionalAttachmentPaths are stored in custom_log.zip.drill-down analysis
In the crash Issue instance, drill-down information for the Issue can be viewed on the drill-down analysis page.
Data Overview
RUM Pro provides data on multiple dimensions for crash Issue analysis, including occurrences, Affected users, Number of affected devices, First Reported Time, and Last reported time.
upward trend
The upward trend page displays statistical charts of Issue reporting trends at different granularities. Granularity options include day, hour, minute, and 5 minutes. Reporting trends can be shown by Occurrences, Affected Users, or Number of affected devices.
Statistical distribution
The analysis supports examining Issue problems through pie charts based on different dimensions (such as application version, OS version, model, etc.). The analysis also allows selection by occurrence count, number of affected users, or number of affected devices.
If you need to analyze the distribution proportion of different device models or architectures under this Issue, or examine the proportion of this problem across different application versions, you can use this statistical distribution feature to view the data.
Interval Distribution
Supports analyzing Issue problems in the form of bar charts based on different fields (such as usage duration, available memory, number of device exceptions, number of user exceptions) and different dimensions (such as occurrence count, number of affected users, number of affected devices). Click Add distribution in the upper right corner to add a new distribution. In the add distribution section, supports inputting intervals such as [0,5), [5,30), [30,60).
error stack Analysis
Supports analyzing issues based on error stack content. Click Add Condition in the upper right corner to add/edit conditions. In the edit condition section, you can add strings to search for, such as searching for the number of individual cases containing the
com string in the current Issue. The search results are displayed comparatively in bar charts.
Help and Support
Was this page helpful?
You can also Contact sales or Submit a Ticket for help.
Help us improve! Rate your documentation experience in 5 mins.
Feedback