KJS crush.

Hi Evgueni,

I think your problem is due to less memory space available (see the

line "Swap space nearly exhausted"), I think your system is running

out of space and it is unhadled xalloc exception.

Try to see the memory requirements of your application (try prstat for

Solaris and perfmon for NT/2000).

Please try to increase your swap space and main memory, and tune your

iAS for these things -

- Heap Allocation (increase the pre-configured heap size appropriate

to the requirements of the application(s) running on the application

server) change

JAVA_AGRS to -Xms256m -Xmx256m to change the heap size to 256 MB.

(For Sp2 change it in ias/bin/kjs script, for Sp3 change it in

ias/env/iasenv.ksh file)

- Garbage Collection Tuning (See

http://java.sun.com/docs/hotspot/gc/index.html for more details about

GC tuning),

- Enable memory optimization for Solaris ( modify

..../ias/env/iasenv.ksh or kjs script to set MEMOPT_KXS=YES) for KJS

and KXS engines,

I think this may help you, for more information please go to this link

-

http://developer.iplanet.com/viewsource/char_tuningias/index.jsp

Please feel free to mail me if you face any problems further.

Sanjeev,

Developer Support Group iAS-India.

Evgueni Gorelenko wrote:

> I have kjs crush problem on sp3, which appear on both NT and Sun

> Solaris platforms.Application includes above 100 servlets and 100

> ejbs.I have put a piece of code, where this problem occur and piece

> of kjs log file

> below.code:-- -try

> { // getting enumeration of ALL win entities in AFSL combo

> allWinsForAFLS = winIF.findByApplication( appNm );while(

> allWinsForAFLS.hasMoreElements() ) {// first get the matching

> entity

>aWin =

> (UCContentWindowEntity)-- -

> kjs: --Unifi

> UnCompile request from remote address: 10.1.6.66. Remote host:

> 10.1.3.105:81

> Swap space nearly exhausted

> SIGSEGV 11 segmentation violation

>si_signo [11]: SEGV

>si_errno [0]:

>si_code [1]: SEGV_MAPERR [addr: 0x0]

> stackpointer=D859E740

> "Uncompile:Default:All:All:en_CA:TEST" (TID:0x88ab7c8,

> sys_thread_t:0x88ab700, state:R, thread_t: t@54,

> threadID:0xd85a0000, stack_bottom:0xd85a0000, stack_size:0x20000)

> prio=5 *current thread*[1]

> com.kivasoft.ebfp.FPRequest.getPayloadObjnative(Native Method)

> [2] com.kivasoft.ebfp.FPRequest.getPayloadObj(Unknown Source)

> [3]

> com.console.entities.ejb_kcp_stub_UCContentWindowEntityHome.findByApplication(e jb_kcp_stub_UCContentWindowEntityHome.java:464)

>

> [4]

> com.console.entities.ejb_stub_UCContentWindowEntityHome.findByApplication(ejb_s tub_UCContentWindowEntityHome.java:103)

>

> [5] com.console.dal.UCDAL.getWinsForAFLS(UCDAL.java:593)

> [6] com.console.compiler.UCWins.getWinsForAFLS(UCWins.java:127)

> [7] com.console.compiler.UCWins.setUncompiledEx(UCWins.java:424)

> [8]

> com.console.compiler.UCUncompile_3x.perform(UCUncompile_3x.java:216)

>

> [9] com.console.compiler.UCUncompile_3x.run(UCUncompile_3x.java:260)

>

> -

> "Thread-45" (TID:0x6b48ab0, sys_thread_t:0x6b489e8, state:CW,

> thread_t: t@52, threadID:0xd8650000, stack_bottom:0xd8650000,

> stack_size:0x20000) prio=1 [1] java.lang.Thread.sleep(Native Method)

>

> [2]

> com.dwl.unifi.servlet.console.Console_AllStates.run(Console_AllStates.java:418)

>

> [3]

> java.lang.Thread.run(Thread.java:478) --

> Thanks in advance. Evgueni.

--

I have never learn not to learn...

at 2007-6-29 9:12:59 >