ape inline push stops working "ape-jsf/Demos/Controller/demo.html" gives "array "

[Mantux]

i was using ape on old server for longet time, php 5.2 debian LAMP
with ssl via stunnel 4.22; now seteled up to new server php 5.3 debian
LAMP 6 squeeze,

always had ape inline push stop to working without giving errors in
log, on the old mashine it wasnt so ferequently,
on new 1 allmoust every night
Tried variety of setups : mpm-itk/m
pm-prefork, APC/xcache, compiled latest build of ape from git source
without errors(clean every time recloning git if error)

searched ape-project group and others(google etc.) all i found there
was problem with ape and "§;%"

mby some1 can share his expierence with me?

[Mantux]

found post familiar to my problem -
https://groups.google.com/group/ape-project/browse_thread/thread/b53e40203710d46b?hl=de

my server spec :

x64 12x1.8ghz i7 6-HPtreaded cores 24GB RAM ocz 280MBs/rw SSD
running Debian 6 (sqeeze) APE 1.1.0 From GIT Source updated on 1.0
core.

Usual work day Connections ~ 500sec - ~100sec (this is an JavaSocket
+LAMP server), newer seen ape going ower 0.1% RAM/CPU load :P
MYSQL 5.5.9; MPM Prefork; Apache2.2.1;

will try test Script from article and post results later on this week.

[Mantux]

ape was running iddle all night, today morning i had same crash
(server version "Server 1.01dev"):
->
[{"time":"1302602227","raw":"ERR","data":
{"code":"001","value":"BAD_PARAMS"}}]
/ape-jsf/Demos/Controller/test.php "line 25"
...
var_dump($data);
$data = json_decode($data);
...

gives :" NULL Error sending message, server response is : "

ape config:

uid {
# "aped" switch to this user/group if it run as root
user = daemon
group = daemon
}


Server {
port = 6969
daemon = yes
ip_listen = x.x.x.x
domain = mydomain.com
rlimit_nofile = 10000
pid_file = /var/run/aped.pid
}

Log {
debug = 1
use_syslog = 0
logfile = ./ape.log
}


JSONP {
eval_func = Ape.transport.read
allowed = 1
}



Config {
#relative to ape.conf
modules = ../modules/lib/
modules_conf = ../modules/conf/
}