Callee not found

, ,
1

Hi first i would try it here,

we call a Natural Programm from Java and get more and more a ‘Callee not found’ Error. Broker is up and Running. Natural Services (3 Running Natural Threads) is Registerd to Broker and other Services wouldn’t give an Error. The called Natural Programm is at the called Library. Sometimes a fresh generated Testprogram (A IDL generated from Natural IDL-Generater using the Workbench; Generate and run Test) would heal the Error and then the Service would answer normaly for a time but then one after one of the 3 Running Naturalserverprocesses would return the ‘Callee not found’ Problem again.

Natural is 3.1.6 - BS2000, Broker is 7.1.2 - BS2000, Client is a Javaprogramm (Java 1.4.2) using EntireX Runtime 7.1.1.60.

Error Message is:

Callee not found. Library: ENTIREX, Program: SLISTSRY, (NATURAL Error Info: Program=SLISTSRY,Error=0082,Line=9999,Status=O,Level=01)

Does anyone have a hint to what i have to change or would it better to open a Request at Servline24?

Greetings Sascha

/Edit: We always have to use conversational Mode becauce of the Size of the Parameters the Natural Programm gets back. And we always use NaturalLogon. (setNaturalLogon, setRPCUserId and setPRCPassword). We don’t use EntireX Security but Natural Security

2

Hi Sascha
Have you checked the basics like:

  1. Steplib of the RPC-server
  2. That the subprog has been stowed
  3. That the FUSER for the PRC-server is the same as the dev-env

Finn

3

As i tried to say sometimes the Program does what i want and sometimes i get an callee not found without changing the Environment.

1.) SYSIDL, SYSRPC, SYSEXT, SYSLIBS, SYSTEM are the Steplibs
2.) If it would’nt be stowed then i always get an error i think :wink:
3.) ???

4

Hi Sascha,
Sorry for misreading ! (and forget my point nr 3!)

  • It appears that my old eyes are failing me :frowning:

I guess the most likely problem is steplib.
I found a support request where one of the subprogam called by the RPC-server did logon to other libs.

Perhaps you could try to add the lib where you subprogs are located to the steplib chain of the RPC-server?

Finn

5

The Service would start at the Lib ENTIREX where also my Program is. The Steblibs of this Lib are that i wrote in my last Posting.

6

Hi Sascha,

What about adding ENTIREX to the steplib of the RPC-server ?

If this doesn’t do the trick perhaps you should make a support request for this to clear up all the technical details versions/other products (like Natural security etc).

Finn

7

Did i understood that right? (I Started profile at the RPC-Server Startlib:)


                                                                                
 15:44:46                    *** NATURAL SECURITY ***                 2006-03-30
                               - Security Profile -                             
                                                                                
 User                                                                           
   ID ....... WIEGANDT                                Link ID ..                
   Name .....                                         ETID .....                
   Type ..... Person                                                            
                                                                                
 Library                                              ----- Steplibs -----      
   ID ....... ENTIREX                                 Library  DBID  FNR        
   Name .....                                         SYSIDL      11   140      
                                                      SYSRPC      11   140      
 Transactions                                         SYSEXT      11   140      
   Startup .. MAINMENU                                SYSLIBS     11   140      
   Restart ..                                         TA1         11   141      
   Error ....                                         TVERSIS     11   141      
                                                      ENTIREX     11   141      
                                                                                
                                                      SYSTEM   ( *STEPLIB )     
 Additional Options ... N                                                       
                                                                                
 Enter-PF1---PF2---PF3---PF4---PF5---PF6---PF7---PF8---PF9---PF10--PF11--PF12---
                   Exit  AddOp Util                                      Canc
8

Yep that was my idea !

BTW you should/could use a different setup for your production RPC.
For instance you won’t need the SYSIDL there (unless you access your sources from prod :wink:
Finn

9

We don’t use Sources at Prod but good to know :smiley: I’ll have a look if the next days the Error comes again. But you have to agree this Solution sounds crazy :wink:

10

  1. have look in broker attribute file, check timeouts for conversations and server non activity

  2. check, that your Natural rpc server is not confugured also as a client :wink: (AUTORPC must not be ON)

  3. keep in mind, that in conversational rpc (after client opened conversation) the rpc server is listening only THIS client and will not
    get freed if client does not call “close conversation” at the end…

11

  1. Both 1 Minute.

  2. Can’t find it at the Startup Procedure. So i think it’s default

  3. I always setup Parameters, open Conversation, make Call, close Conversation and disconnect from Broker.

I’ll open an Support Request at Servline :roll: with Link to this Thread.

12

My suggestions:

  1. Use SMH to identify all server replicates which have registered to the same server name. Check their Natural configurations, maybe there is a bad guy with a wrong configuration.

  2. Run the Natural RPC Servers with trace enabled to see what’s going on when the error occurs.

13

Hi Sacha,
We had a similar problem with the following scenario:

  • Server issues LOGON
  • Each client request also does an explicit LOGON to this same
    After we changed server LOGON to LOGON SYSTEM we didn’t get no more NAT0082.

Aldo Teusch

14

  1. SMH is only running on non-BS2000 Systems AFAIK. So i use Broker Controll Center but i get displayed all RPC-Servers normaly.

  2. I belive i should send such a trace at the next days to my Support Request :smiley: (Few Minutes ago i had to send a Javatrace)

@Aldo we have to logon to the User-called Library. We wrote a Wrapper for USR1057N so we could get Sources to Java and use External Tools like WinMerge to show the Developer a good Programmdiff. And some more like that.