Hi Rudi,
Thanks for the your inputs. Please find my comments below inline for your queries.
Regards
Abhishek
------- Original Message -------
Sender : Rudolf Streif<[email protected]>
Date : Apr 16, 2014 00:22 (GMT+09:00)
Title : Re: Proposal for new component in Tizen IVI
1. How would you compare your Health Manager to GENIVI's Diagnostic Log and Trace (DLT) [1]? What are the commonalities, what are the differences? DLT aims to be Autosar 4.0 compliant. Is that a goal for the HM too?
<Abhishek> As per my understanding DLT is an Autosar component responsible for the logging and tracing functionality for developement purpose, specifically for ECUs running AUTOSAR 4.0. It offers similar functionalities but certainly not limited to DEM (Diagnostic Event Manager) in terms of storing the fault data. HM stores pertinant fault-data of the services from monitoring and recovery aspects - the details about those could be found in the ppt shared with the group.
2. Since you are building on dbus and systemd, have you considered journald as logging mechanism? I think it would be beneficial having all system logs in one database.
<Abhishek> As mentioned in HM, we are not looking for log messages of the service rather we are looking at monitoring the critical service, and capturing the fault-data. We also proposed a placeholder for applying predefined OEM specific recovery strategy for diagnosed faults.
3. What security mechanisms are you putting in place to protect the logs?
<Abhishek> It is an important input/suggestion. Since the proposal is in initial stage, we can definately discuss and take it forward while developing requirements doc for this.
4. What mechanisms is HM providing for remote access to the logs?
<Abhishek> Our proposition is that captured data will be used for onboard diagnosis and planned recovery. We are also envisioning that the logged data, if a requirement arises so, could be copied to a Gateway device from it could be remotely accessed.
5. The block diagram includes a box labeled "Automotive Peripheral Framework (CAN/LIN/MOST)". Are you intending to use the Automotive Message Broker (AMB) of Tizen IVI for that functionality?
<Abhishek> Yes, that is the intent. We can use the AMB of Tizen IVI as well in future.
6. When you are referring to "services" that the HM monitors are these daemons and processes that are managed using systemd?
<Abhishek> Yes, these (daemons and processes) are the critical services registered with HM and managed by systemd.
7. Can any service be monitored by HM or does a service have to be prepped for that? The block diagram seems to suggest that the services have to implement specific APIs and expose them via the Session DBus.
<Abhishek> The HM shall monitor the predefined critical services but it also exposes APIs on dbus session bus where other services which want to be monitored can register themselves and query their status and take corrective actions themselves like Function inhibition.
Dear All,
We would like to propose a new component called System Health Manager for Tizen IVI. It is imperative that automotive platform should provide a means to monitor critical services and take corrective action whenever a failure or deviation from expected services is detected. In that respect, the proposed Health Manager (HM) shall seamlessly integrate with Tizen IVI for providing a limited fault-tolerance capability. Attached please find the presentation slides which describe overall functioning and features of the proposed HM. We request you to provide your comments/feedback on our proposal, and ways for taking this initiative forward.
The key features of the proposed HM include detecting disruption of services, logging data required for diagnosis and applying directed recovery strategies.The proposed HM uses systemd interfaces and signals to monitor configured critical services and recover them, if failed.
The initial development is undergoing and can be released at tizen.org for open source contributions following the feedback and internal due process for releasing the code. At the moment, we have basic structure for HM core, Dbus Communication and Fault-database. We have also designed a placeholder for dummy recovery client at the moment. Going forward this needs to be developed as per system requirements.
Looking forward to your suggestions, feedback and active contributions of this feature on tizen.org for IVI profile.
Regards
Abhishek
_______________________________________________
IVI mailing list
[email protected]
https://lists.tizen.org/listinfo/ivi
Rudolf J. Streif
Linux Foundation Training Schedule: training.linuxfoundation.org
|
|
_______________________________________________ IVI mailing list [email protected] https://lists.tizen.org/listinfo/ivi
