Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
DynSize and 5333
127 views
Skip to first unread message
Martin
May 29, 2012, 4:21:30 AM5/29/12
to
Hello
Our 2.7 app had it set to 256, our 2.8 WAS 128 but now 256
However we are getting 5333 with dynsize 256 which the dynsize 128 doesn't.
Any ideas what is happening please?
Thanks
Martin
Our 2.7 app had it set to 256, our 2.8 WAS 128 but now 256
However we are getting 5333 with dynsize 256 which the dynsize 128 doesn't.
Any ideas what is happening please?
Thanks
Martin
Paul D B
May 29, 2012, 10:13:48 AM5/29/12
to
256? In my apps (but they aren't "huge") I've got it only set to 40.
That's 40 pages of 64 KB. The apps run OK that way so I don't touch it.
I recall that in some apps with heavy graphics, If I set it higher, I run
into problem
"31 Kid stack Limit Exceeded".
Which is caused by more than 16000 unreferenced objects. I could increase
that limit with SetMaxRegisteredAxitMethods() but I never bothered.
Anyway, what I think: the higher you set the Dynamic memory the less often
the GC kicks in and that might be a reason why you run into 5333 ?
PS: You could, in the method that's crashing, insert a CollectForced().
Worth a try.
Paul
"Martin" schreef in bericht news:aG%wr.1002879$ed1.8...@fx04.am4...
That's 40 pages of 64 KB. The apps run OK that way so I don't touch it.
I recall that in some apps with heavy graphics, If I set it higher, I run
into problem
"31 Kid stack Limit Exceeded".
Which is caused by more than 16000 unreferenced objects. I could increase
that limit with SetMaxRegisteredAxitMethods() but I never bothered.
Anyway, what I think: the higher you set the Dynamic memory the less often
the GC kicks in and that might be a reason why you run into 5333 ?
PS: You could, in the method that's crashing, insert a CollectForced().
Worth a try.
Paul
"Martin" schreef in bericht news:aG%wr.1002879$ed1.8...@fx04.am4...
D.J.W. van Kooten
May 29, 2012, 11:12:07 AM5/29/12
to
On Tue, 29 May 2012 16:13:48 +0200, "Paul D B"
<pol...@NOMORESPAMhnt.be> wrote:
Hello Paul, Martin,
>256? In my apps (but they aren't "huge") I've got it only set to 40.
Hmmm.. I've set it to 512 without specific problems. Read 5333 errors
do not happen that often so they are hard to solve (as usually not
reproducable). The cause is NOT Dynsize but something else you should
try to solve, although I agree with Paul that this setting may
influence the occurance of the 5333 due to the GC kicking in. But also
with your current setting I suspect the 5333 will come up sooner or
later if you don't find the cause.
Dick
<pol...@NOMORESPAMhnt.be> wrote:
Hello Paul, Martin,
>256? In my apps (but they aren't "huge") I've got it only set to 40.
do not happen that often so they are hard to solve (as usually not
reproducable). The cause is NOT Dynsize but something else you should
try to solve, although I agree with Paul that this setting may
influence the occurance of the 5333 due to the GC kicking in. But also
with your current setting I suspect the 5333 will come up sooner or
later if you don't find the cause.
Dick
Paul D B
May 30, 2012, 4:06:55 AM5/30/12
to
It all depends on the application. As I said, the higher you set it, the
more unreferenced objects remain in memory (because the GC only kicks in
less often ) and there are limits to that too (16000 objects by default
IIRC). But in most applications you will never have that amount of objects
so you will never run into problems by setting Dynsize very high.
I agree that playing with Dynsize or doing Collectforced() to force the
garbage collector, are not real solutions (the cause is elsewhere, as you
said) but it might help, and may "hide" the sporadic 5333 problem :-)
Paul
"D.J.W. van Kooten" schreef in bericht
news:ckp9s714abki807mq...@4ax.com...
more unreferenced objects remain in memory (because the GC only kicks in
less often ) and there are limits to that too (16000 objects by default
IIRC). But in most applications you will never have that amount of objects
so you will never run into problems by setting Dynsize very high.
I agree that playing with Dynsize or doing Collectforced() to force the
garbage collector, are not real solutions (the cause is elsewhere, as you
said) but it might help, and may "hide" the sporadic 5333 problem :-)
Paul
"D.J.W. van Kooten" schreef in bericht
news:ckp9s714abki807mq...@4ax.com...
0 new messages