Maybe not everyone watched the discussion we had about this. The point is that we have
to differenctiate between pool and cache. In the pool are beans that are not assigned
to a specific ID. In the cache, there ist the pool plus formerly used (and with this,
loaded) beans that ARE assigned to a specific ID. PoolMin is used to preallocated
empty beans for making JOnAS faster at startup because it is likely that JOnAS will
need to allocate many beans in the next time. CacheMax is the upper limit of all
beans, with and that without ID, that JOnAS will keep in memory; it is needed for not
running out of memory; you limit cache size and with this you can keep a security
amount of memory (the space between memory size and (cachemax * size_per_bean +
overhead)) for safe and stable work of the server.
> Christophe Ney wrote:
> >
> > I like the new names,
> > I understand that there is no need for a min-cache-size,
> > But why don't we have a max-pool-size ?
> > christophe
> The need is for scalability, to limit the number of instances in memory.
> max-cache-size is used for this.
> Why should we have a max-pool-size then ?
> We could also name these values:
> max-number-of-instances-in-memory and min-number-of-instances-in-memory
> :-)
> The goal was to have short names: poolmin and cachemax but may be it's
> not a good idea.
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
----
To unsubscribe, send email to [EMAIL PROTECTED] and
include in the body of the message "unsubscribe jonas-users".
For general help, send email to [EMAIL PROTECTED] and
include in the body of the message "help".
