understanding the gc log
Hi,
I'm trying to understand the garbage collection log (1.4.2). I have log statements in there like:
14325.110: [GC 14325.110: [DefNew: 111691K->111691K(118016K), 0.0000206 secs]14325.111: [Tenured: 283068K->298347K(393216K), 4.1681840 secs] 394760K->298347K(511232K), 4.1683295 secs]
The way I read it is: nothing done on the New Generation and the tenured Generation even increased. But looking at the total heap, the total used memory after the garbage collection equals the total used memory in the tenured Generation making me believe that the New Generation is empty. Is there a display bug that the New Generation was completely cleared but only the log incorrect? But is it really possible to clear the New Generation up to the very last byte?
dirk
[798 byte] By [
furballa] at [2007-10-3 0:45:03]
> Hi,
>
> I'm trying to understand the garbage collection log
> (1.4.2). I have log statements in there like:
>
> 14325.110: [GC 14325.110: [DefNew:
> 111691K->111691K(118016K), 0.0000206 secs]14325.111:
> [Tenured: 283068K->298347K(393216K), 4.1681840 secs]
> 394760K->298347K(511232K), 4.1683295 secs]
>
> The way I read it is: nothing done on the New
> Generation and the tenured Generation even increased.
> But looking at the total heap, the total used memory
> after the garbage collection equals the total used
> memory in the tenured Generation making me believe
> that the New Generation is empty. Is there a display
> bug that the New Generation was completely cleared
> but only the log incorrect?
Yes, exactly.
> But is it really possible
> to clear the New Generation up to the very last
> byte?
Yes, all the survivors in the New Generation were compacted
up into the Old Generation. (The whole heap was collected.)
