Including kernel object information in a user dump

Patent metadata

Field	Value
Publication number	US-10114731-B2
Application number	US-201414581392-A
Country	US
Kind code	B2
Filing date	Dec 23, 2014
Priority date	Dec 30, 2013
Publication date	Oct 30, 2018
Grant date	Oct 30, 2018

Abstract

An improved method of analyzing software issues may include retrieving and storing selected data elements from the operating system kernel data prior to performing a memory dump. The method of retrieving the selected kernel data may include creating a thread dedicated to collecting the data and storing it in a memory location for analysis after the memory dump. The operating system kernel data may be analyzed in conjunction with the prior art dump data to identify a root cause of the software issue.

First claim

What is claimed is:

1. A method of identifying a software issue in a data storage system, comprising:
storing selected operating system kernel data in a memory location of the data storage system;
analyzing the stored operating system kernel data in the memory location identifying a root cause of the software issue from the analyzed operating system kernel data; and
transmitting an alert to a user after the root cause is identified, the alert identifying the root cause to the user;
wherein the method further comprises:
performing, after storing the selected operating system kernel data in the memory location, a memory dump operation configured to output memory dump data of a software process associated with the data storage system;
storing memory dump data in another memory location; and
analyzing the memory dump data, analysis of the memory dump data and the operating system kernel data providing the identification of the root cause of the software issue;

wherein the selected operating system kernel data is stored at a time selected based on a number of available handles in a thread of the data storage system; and
wherein the method further comprises:
comparing the number of available handles to a selected minimum handle threshold level of available handles for the thread;
when the number of available handles is below the selected minimum handle threshold level, generating a collection thread to collect file name data for each file handle from the operating system kernel;
storing the file name data in the memory location; and
initiating the memory dump operation.

2. The method as in claim 1 , further comprising:
integrating the stored selected operating system kernel data in the memory location with the stored data resulting from the memory dump operation in the other memory location to form an integrated data set; and
analyzing the integrated data to provide identification of the root cause of the software issue.

3. The method as in claim 2 , further comprising;
generating, after the root cause is identified, a suggested solution to the software issue; and
transmitting the suggested solution to the user.

4. The method as in claim 1 , wherein storing the selected operating system kernel data further includes:
generating at least one thread for gathering sync object data, the sync object data including at least one of mutex, semaphore, event, critical section, process thread hanging and critical timeout data; and
storing the gathered sync object data in the memory location.

5. The method as in claim 4 further comprising:
obtaining, when gathering critical timeout data, an address of a different software process associated with the critical timeout data;
generating a memory dump of selected data from the different software process;
storing the memory dump of selected data from the different software process in a memory location;
analyzing the operating system kernel data in conjunction with the memory dump of selected data of the different software process to identify a root cause of the software issue; and
transmitting an alert to the user after the root cause is identified.

6. The method as in claim 5 further comprising:
generating a suggested solution to the software issue; and
transmitting the suggested solution to at least one of the user and a second user of the different software process.

7. The method of claim 1 , wherein the method further includes, in response to an instruction to delete a file:
storing a file name associated with the file and a file handle associated with the file at a time prior to receiving the instruction;
detecting, after the file has been deleted, that the file handle remains associated with the file; and
in response to detecting that the file handle remains associated with the file, including the file name and the file handle in the selected operating system kernel data.

8. The method of claim 1 , wherein the method further comprises including file handle trend data in the selected operating system kernel data, the trend data including a history of a number of available file handles in the thread of the data storage system;
wherein analyzing the selected operating system kernel data includes detecting a trend in the file handle trend data which indicates a handle leak software problem exists and identifying it as the root cause of the software issue; and
wherein transmitting the alert includes identifying file handle leakage as the root cause of the software issue to the user.

9. A system for identifying a software issue in a computer, comprising:
a communication interface;
a memory;
a processing circuit, including a controller, coupled to the memory and communication interface, the controller being constructed and arranged to:
execute, by the processing circuit, a series of computer commands to perform a software process;
store selected operating system kernel data in a first memory location of the memory;
analyze, by the processing circuit, the operating system kernel data in the first memory location to identify a root cause of the software issue; and
transmit, by the communication interface, an alert to a user when a root cause is identified;

wherein further the controller is constructed and arranged to:
perform, after storing the selected operating system kernel data in the first memory location, a memory dump operation of a software process associated with the data storage system;
store memory dump data resulting from the memory dump operation in a second memory location;
analyze, by the processing circuit, the data in the first and second memory locations to identify a root cause of the software issue; and

wherein further the controller is constructed and arranged to:
store the selected operating system kernel data at a time selected by measuring, by the processing circuit, a number of available handles in a thread of the data storage system;
compare, by the processing circuit, the number of available handles to a selected minimum handle threshold level of available handles for the thread;
create, when the number of available handles is below the selected minimum handle threshold level, a thread to collect file name data for each file handle from the operating system kernel;
store the file name data in the first memory location; and
initiate the memory dump operation.

10. The system as in claim 9 , wherein further the controller is constructed and arranged to:
integrate the stored selected operating system kernel data in the first memory location with the stored memory dump data in the second memory location;
generate, when the root cause is identified, a suggested solution to the software issue, and
transmit the suggested solution to the user.

11. The system as in claim 9 , wherein the controller is constructed and arranged to store the selected operating system kernel data, further includes:
create at least one thread for gathering sync object data, the sync object data including at least one of mutex, semaphore, event, critical section, process thread hanging and critical timeout data; and
store the gathered sync object data in the first memory location.

12. The system as in claim 11 wherein further the controller is constructed and arranged to:
obtain, when gathering critical timeout data, an address of a different software process associated with the critical timeout data;
generate a memory dump of selected data from the different software process;
store the memory dump of selected data in a third memory location; analyze the operating system kernel data in the first memory location in conjunction with the memory dump of selected data of

Assignees

• Emc Corp
• Emc Ip Holding Co Llc

Inventors

• Fan Xiaohua
• Yin Feng
• Wang Xiaogang
• Dong Dazhi
• Lu Binhua

Classifications

• G06F11/0715

  in a system implementing multitasking (multitasking per se G06F9/46) · CPC title

• G06F11/366 Primary

  using diagnostics (G06F11/0703 takes precedence) · CPC title

• G06F11/0778 Primary

  Dumping, i.e. gathering error/state information after a fault for later diagnosis · CPC title

• G06F11/079

  Root cause analysis, i.e. error or fault diagnosis (in a hardware test environment G06F11/22; in a software test environment G06F11/36) · CPC title