Scalable logging control for distributed network devices

• US 9,807,154 B2
• Filed: 09/26/2014
• Issued: 10/31/2017
• Est. Priority Date: 09/26/2014
• Status: Active Grant

• Current Assignee: Lenovo Global Technologies International Ltd. (Lenovo Group Ltd.)
• Original Assignee: Lenovo Enterprise Solutions (Singapore) Pte., Ltd. (Lenovo Group Ltd.)
• Inventors: Darling, Glen, Jacob Da Silva, Roberto H.
• Primary Examiner(s): Nguyen, Quang N
• Application Number:
  US14/498,937
• Publication Number:
  US 20160094620A1
• Time in Patent Office: 1,131 Days
• Field of Search: 709203, 709204, 709206, 709217, 709219, 709223, 709224, 709226, 709228, 709231, 709232, 709238, 714 15, 725107
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

In one embodiment, a method includes receiving one or more logging events from a client application during a period of time when a logging control configuration has not been fully applied, handling any logging events directed toward at least one of a library code and one or more client applications by storing information relating to an originator of the logging events in a log file and dispatching the logging events directed toward at least one of the library code and the one or more client applications to a syslog daemon, and en-queuing all other logging events for later processing in a circular buffer maintained in non-persistent memory of a shared memory for later retrieval, wherein the other logging events are en-queued for later processing regardless of a severity indicated for any individual logging events. Other methods, systems, and computer program products are described according to more embodiments.

Claims

1. A system, comprising:

a master node controller comprising a processor and logic integrated with and/or executable by the processor, the logic being configured to;

receive one or more logging events from a client application during a period of time when a logging control configuration of a logging control application has not been fully applied;

handle any logging events directed toward at least one of a library code and one or more client applications by storing information relating to an originator of the logging events in a log file and dispatching the logging events directed toward at least one of the library code and the one or more client applications to a syslog daemon;

en-queue all other logging events independent of a severity rating for any individual logging events for later processing in a shared memory, wherein the shared memory is owned by the logging control application; and

processing the en-queued logging events in response to determining that the logging control configuration has been fully applied.

2. The system as recited in claim 1, wherein the other logging events are stored to a circular buffer maintained in non-persistent memory of the shared memory for later retrieval.

3. The system as recited in claim 2, wherein the logic is configured to:

detect that buffer overflow has occurred causing some of the other logging events to be discarded; and

log a warning message after processing the other logging events from the circular buffer, the warning message specifying a number of logging events which have been discarded.

4. The system as recited in claim 2, wherein the logic is configured to:

retrieve the other logging events from the circular buffer for processing after the logging control configuration has been fully applied, as indicated by a logging service ready state being enabled;

filter the other logging events based on a severity of individual logging events to provide filtered logging events after the logging control configuration has been fully applied; and

replay the filtered logging events by sending the filtered logging events to the syslog daemon for handling thereof after the logging control configuration has been fully applied.

5. The system as recited in claim 2, wherein the logic is configured to:

retrieve the other logging events from the circular buffer when the logging control configuration has not been fully applied, as indicated by a logging service ready state not being enabled; and

replay the other logging events by sending the other logging events to the syslog daemon for handling thereof, regardless of a severity indicated for any individual logging events, prior to the logging control configuration being fully applied.

6. The system as recited in claim 2, wherein the logic is further configured to create and/or allocate the shared memory using at least one of the logging control application and the one or more client applications prior to storing logging events thereto, wherein the shared memory is accessible to the logging control application and the one or more client applications, and wherein the shared memory is owned by the logging control application.

7. The system as recited in claim 6, wherein a client application triggers creation and/or allocation of the shared memory after determining that the shared memory has not already been created and/or allocated.

8. The system as recited in claim 1, wherein the logic is configured to determine whether a global flag indicating a service status of a log agent for at least one of the client applications is set to true,

wherein the global flag is readable via an application programming interface (API), and

wherein the global flag is set to true only when:

the shared memory is created and/or allocated for the log agent; and

a networking software start-up configuration is parsed and provided to a log manager.

9. The system as recited in claim 8, wherein the logic is further configured to process logging events provided by any client application for which the global flag of a corresponding log agent is set to true by sending the logging events to the syslog daemon.

10. The system as recited in claim 8, wherein the logic is further configured to add all logging events in the circular buffer to a core dump file related to a node controller which issued the logging events when a fatal exception occurs before the global flag is set to true.

11. The system as recited in claim 1, wherein the logic configured to en-queue all other logging events for later processing comprises:

logic configured to write a final string and syslog message string of each logging event to the shared memory in a portion of the shared memory allocated to a log agent instance for which the logging events pertain.

12. A method, comprising:

receiving one or more logging events from a client application during a period of time when a logging control configuration of a logging control application has not been fully applied;

handling any logging events directed toward at least one of a library code and one or more client applications by storing information relating to an originator of the logging events in a log file and dispatching the logging events directed toward at least one of the library code and the one or more client applications to a syslog daemon; and

during the period of time when the logging control configuration of the logging control application has not been fully applied, en-queuing all other logging events for later processing in a circular buffer maintained in non-persistent memory of a shared memory for later retrieval,

wherein the other logging events are en-queued for later processing regardless of a severity indicated for any individual logging events.

13. The method as recited in claim 12, further comprising:

detecting that buffer overflow has occurred causing some of the other logging events to be discarded;

logging a warning message after processing the other logging events from the circular buffer, the warning message specifying a number of logging events which have been discarded;

retrieving the other logging events from the circular buffer for processing after the logging control configuration has been fully applied, as indicated by a logging service ready state being enabled;

filtering the other logging events based on a severity of individual logging events to provide filtered logging events after the logging control configuration has been fully applied; and

replaying the filtered logging events by sending the filtered logging events to the syslog daemon for handling thereof after the logging control configuration has been fully applied.

14. The method as recited in claim 12, further comprising:

creating and/or allocating the shared memory using at least one of the logging control application and the one or more client applications prior to storing logging events thereto, wherein the shared memory is accessible to the logging control application and the one or more client applications, and wherein the shared memory is owned by the logging control application, wherein a client application triggers creation and/or allocation of the shared memory after determining that the shared memory has not already been created and/or allocated;

retrieving the other logging events from the circular buffer when the logging control configuration has not been fully applied, as indicated by a logging service ready state not being enabled; and

replaying the other logging events by sending the other logging events to the syslog daemon for handling thereof, regardless of a severity indicated for any individual logging events, prior to the logging control configuration being fully applied.

15. The method as recited in claim 12, further comprising determining whether a global flag indicating a service status of a log agent for at least one of the client applications is set to true,

wherein the global flag is readable via an application programming interface (API), and

wherein the global flag is set to true only when:

the shared memory is created and/or allocated for the log agent; and

a networking software start-up configuration is parsed and provided to a log manager.

16. The method as recited in claim 15, further comprising:

processing logging events provided by any client application for which the global flag of a corresponding log agent is set to true by sending the logging events to the syslog daemon; and

adding all logging events in the circular buffer to a core dump file related to a node controller which issued the logging events when a fatal exception occurs before the global flag is set to true.

17. The method as recited in claim 12, wherein the en-queuing all other logging events for later processing comprises writing a final string and syslog message string of each logging event to the shared memory in a portion of the shared memory allocated to a log agent instance for which the logging events pertain.

18. A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions being executable by a master node controller to cause the master node controller to:

receive one or more logging events from a client application during a period of time when a logging control configuration of a logging control application has not been fully applied;

handle any logging events directed toward at least one of a library code and one or more client applications by storing information relating to an originator of the logging events in a log file and dispatching the logging events directed toward at least one of the library code and the one or more client applications to a syslog daemon;

create and/or allocate a circular buffer maintained in non-persistent memory of a shared memory using at least one of the logging control application and the one or more client applications, wherein the shared memory is accessible to the logging control application and the one or more client applications, and wherein the shared memory is owned by the logging control application; and

en-queue all other logging events for later processing in the circular buffer of the shared memory for later retrieval regardless of a severity indicated for any individual logging events.

19. The computer program product as recited in claim 18, wherein the program instructions are further executable by the master node controller to:

detect that buffer overflow has occurred causing some of the other logging events to be discarded;

add all logging events in the circular buffer to a core dump file related to a node controller which issued the logging events when a fatal exception occurs before a global flag is set to true, wherein the global flag is set to true only when;

the shared memory is created and/or allocated for a log agent; and

a networking software start-up configuration is parsed and provided to a log manager;

log a warning message after processing the other logging events from the circular buffer, the warning message specifying a number of logging events which have been discarded;

retrieve the other logging events from the circular buffer for processing after the logging control configuration has been fully applied, as indicated by a logging service ready state being enabled;

filter the other logging events based on a severity of individual logging events to provide filtered logging events after the logging control configuration has been fully applied; and

replay the filtered logging events by sending the filtered logging events to the syslog daemon for handling thereof after the logging control configuration has been fully applied.

20. The computer program product as recited in claim 19, wherein the period of time of logging control configuration is during start up of the master node controller,

wherein the logic is further configured to add all logging events in the circular buffer to a core dump file related to a node controller which issued the logging events when a fatal exception occurs before the global flag is set to true;

and

wherein the en-queuing all other logging events for later processing comprises writing a final string and syslog message string of each logging event to the shared memory in a portion of the shared memory allocated to a log agent instance for which the logging events pertain.