[EMAIL PROTECTED] wrote on 04/22/2008 07:37:50
AM:
> Hello All!
>
> When exploring the latest available version of the OpenHPI SNMP Subagent
> (version 2.3.4), we discovered the following issues:
>
> 1. The SubAgent caches sensor reading for all known sensors and updates
the
> cached reading only on sensor creation and when a sensor event comes.
When a
> User requests sensor reading, he/she is provided with cached value and
no
> query for the up-to-date value is made. If a sensor reading changes
without
> any event generation, the User is provided with an out-of-date value.
>
> 2. The same applies to getting control state and mode. The situation is
even
> worse here since controls do not generate any events and the cached
control
> values are never updated. In addition, when the control representation
is
> created in the Subagent, its cached state and mode is initialized from
the
> default state and mode, not from the actual state and mode.
>
> The question is, were these issues a consequence of a conscious design
> decision made by the Subagent developers not to update the state of
sensor
> readings and control data at runtime? If so, what was the motivation for
> this design decision?
>
I don't believe it was a concious decision.
Honeslty, the snmp-subagent is without a maintainer. I would be great if
you could submit a patch to fix it. I don't think it will ever get fixed
if you don't unless a willing maintainer comes along.
--Renier-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Openhpi-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openhpi-devel
