Server crashing when custom Agent program is run

HCL Domino Domino Forum
1

Hi all,Background - We have a custom Agent program written to sync Contacts, Appointments and To-Dos between user’s Lotus Notes and our CRM application database, both ways.

The Agent is deployed as a .NSF database on the Server and scheduled to run every 1 hour.

The Agent loops through each User mail file, picks up incremental data and creates the same in our application database. Then, updates are picked up from our application database and corresponding documents are created in the user’s mail file.

The Program is signed using the Admin ID from the Server.

Issue -

The Agent program mostly runs fine and does what is expected. But, number of times, what has happened is that the program leads to crashing of the server, before it could complete.

I am new in the area and am not able to get to the root of this.

  • Why is the server crashing if the Agent program is encountering an error scenario?

  • Is it because full permissions to the user mail files are not available to the Admin ID?

  • Is there a way to trace/ log exact cause for the problem?

Your comments are greatly appreciated. Waiting for your replies.

Thanks and Regards,

Sameer

P.S - This is the stack trace from the server log, in case it might be useful.

Build: 8.5FP1 June 15, 2009 (64-bit

Host Name: AIX ios13204e 1 6 0002546AD400

Date: Tue Aug 3 10:17:04 EDT 2010

##################################

thread 14/14 :: amgr pid=8192186, k-id=62783517 , pthr-id=3086

stack :: k-state=wait, stk max-size=264328, cur-size=10616

##################################

raise.nsleep(??, ??) at 0x90000000002b7c4

sleep(??) at 0x90000000014ece8

OSRunExternalScript(0x11357fa40, 0x100000001) at 0x900000000c4ee44

OSFaultCleanupExt(0x0, 0x100000001000, 0x0, 0x0, 0x0, 0x0) at 0x900000000c4e67c

OSFaultCleanup(0x0, 0x100000001000, 0x0) at 0x900000000c50034

fatal_error(0xb0000000b, 0x113580130, 0x11357fe80) at 0x9000000010f0a90

.() at 0x3434

OSLockReadSem(0x1430) at 0x900000000c14f3c

LockDbReadSemCtx(0x113580860, 0x100000001) at 0x900000000f9bedc

LockDbROCond(0x113580860, 0x1a00000000001a, 0x100000001) at 0x900000000f9ab8c

NSFDbGetNoteInfoExtended(0x0, 0x10f6000010f6, 0x0, 0x0, 0x113580990, 0x113580994, 0x0, 0x0) at 0x900000001034e50

ANGetPropBoolean__6ANNoteFUs(0x1136603a0, 0xb8f000000000b8f) at 0x900000005e5ac78

ANGetPropObject__11ANViewEntryFUs(0x11008dde8, 0x7c70000000007c7) at 0x900000005ee9ba4

ANJPropGetAdt__6ANotesFP7JNIEnv_i(0x11008dde8, 0x112840c00, 0x7c7000007c7) at 0x900000005dedde8

Java_lotus_domino_local_NotesBase_PropGetAdt(0x112840c00, 0x1128d52a8, 0x7c7000007c7) at 0x900000005debb18

@REGS

2

Subject: POST YOUR CODE

3

Subject: This is speculation

Without seeing code area that is causing the crash I couldn’t give an accurate answer.

But from the crash stack. You are running a java agent. As you mention you are new to this, then the most likely cause of the crash is due to the incorrect use of recycle() handling.

Please see the following.

http://www-01.ibm.com/support/docview.wss?uid=swg21097861

Also a good write up on it.

… If recycle is not the issue, then I would recommend putting debug lines into the agent. Using the following notes.ini settings you can check the console logs in IBM_TECHNICAL_SUPPORT folder to see if the crash is consistent.

DEBUG_THREADID=1

CONSOLE_LOG_ENABLED=1