On 09/04/07, Niclas Hedhman <[EMAIL PROTECTED]> wrote:

The specification is not clear what should happen if the EventAdmin is not
available. Availability could be temporarily gone (e.g. update) or more
permanently missing.

Three choices as far as I can tell;

 1. Discard all LogEvents.
 2. Buffer all LogEvents and deliver when possible.
 3. Buffer LogEvents for a while, discard the rest.

Any takers on what is the desired functionality in an enterprise setting?


Option 2 could lead to OutOfMemory issues in the enterprise, especially
if a rogue/malfunctioning bundle is repeatedly issuing log events.

Option 1 is much safer, but one of the discarded events could be the one
that explains why the system is acting up.

How about option 3, but using a circular buffer to keep the last N events?
Typically the events you are interested in are in the last few cycles.

( this is the approach we used in the JVM for the internal "flight recorder" )


Cheers
--
Niclas Hedhman, Software Developer

I  live here; http://tinyurl.com/2qq9er
I  work here; http://tinyurl.com/2ymelc
I relax here; http://tinyurl.com/2cgsug
_______________________________________________
OSGi Developer Mail List
[email protected]
http://www2.osgi.org/mailman/listinfo/osgi-dev


--
Cheers, Stuart
_______________________________________________
OSGi Developer Mail List
[email protected]
http://www2.osgi.org/mailman/listinfo/osgi-dev

Reply via email to