Two RPC services can't co-exist without one of them issuing 3009 errors

[Andy_Edwards2 (Andy Edwards)]

Have tried different ADM2 plus ETID values in NatSec to no avail
Both run fine as standalones (without the other). No consistency
in which service will produce the 3009s

ADA854, NAT912, ENTIREX 10.9, NDV912

‘NAT3009 Last transaction backed out of database 78. Subcode 66’ for all transactions

[Finn_the_Dane1 (Finn the Dane)]

Hi Andy - this is a known “feature” - and solved by applying NATRPC39
Finn
See full doc here

Avoiding Error Message NAT3009 from Server Program

If a server application program does not issue a database call during a longer period of time, the next database call might return a NAT3009 error message.

To avoid this problem, the optional user exit NATRPC39 is provided. This exit is called in the following cases:

1. After a Ping command;
2. In case of communication via the EntireX Broker: each time the SERVER-NONACT time is exceeded.

To activate the user exit NATRPC39

1. Copy the sample exit NATRPC39 from library SYSRPC to library SYSTEM on system file FUSER.The steplib concatenation of the library to which the server currently is logged on is not evaluated.
2. Adapt the database ID which is assigned to field ACB-RSP to your needs.
3. Add additional CALL 'CMADA' statements, which reference additional database IDs to your NATRPC39 if more than one Adabas database is involved.

[Andy_Edwards2 (Andy Edwards)]

Apologies, forgot to mention this is already in place. Thanks anyway

[wolfgang.winter]

Interesting, there is no subcode 6 for a Rsp 9 …

What ETID settings have you tried ? F or S should work as they include a timestamp in the actual ETID.

[Andy_Edwards2 (Andy Edwards)]

As requested Wolfgang. Thanks
- Preset User Values -

ETID … U

   ETID

---

Default … *USER

[wolfgang.winter]

So does it work when you change it to an “F” ?

[Andy_Edwards2 (Andy Edwards)]

Thanks for the response.
We dare not use ‘F’ due to it’s similar functionality to ‘S’ which messed up our ETIDs whilst using DPS
Apologies but the Subcode is a 66.
Incidentally, we’re zIIP enabled

[Andy_Edwards2 (Andy Edwards)]

Here’s an update just in case anybody encounters the same issue, a workaround of amending the RPC PROC and adding parameter ‘DBOPEN=ON’ seems to have worked. Any comments why that would work would be appreciated. SAG investigations ongoing

[Leon_Matthee]

Hi Andy ,

What is the path of the RPC PROC?

[Andy_Edwards2 (Andy Edwards)]

Hi Leon, PROC is in SYS3.PROCLIB, is that what u meant?