Article ID: 138469
Article Last Modified on 2/22/2007
APPLIES TO
- Microsoft SNA Server 4.0
- Microsoft SNA Server 4.0 Service Pack 4
- Microsoft Host Integration Server 2000 Standard Edition
- Microsoft Host Integration Server 2000 Service Pack 1
This article was previously published under Q138469
SUMMARY
If you have more than one mainframe attached to your Front End Processor
(FEP) with Network Control Program (NCP) running, the IBM SNA "error
recovery protocol" (ERP) can be used. This can be used for ownership
takeover without interrupting the active LU-LU session.
For example, if the
owning Service Control Point (SSCP) of PU/LU is down, the LU without active
session becomes inactive. The other SSCP/mainframe can take the ownership
of PU/LU without resetting the active LU-LU session by ACTPU(ERP) and ACTLU
(ERP). And the other LU can now be activated by the new SSCP.
SNA Server supports the recovery of PLU-SLU sessions after receiving an
ACTPU (ERP) and ACTLU (ERP) command. However, ACTPU (ERP) recovery requires
that the SNA Server connection is not dropped. SNA Server reports
connection failures by logging Event 23 in the Windows application event
log. If the SNA Server connection to the FEP is lost, PLU-SLU session
recovery is not possible.
REFERENCES
For information about the ACTPU and ACTLU message formats,
see the IBM SNA Formats Guide (IBM p/n# GA27-3136), Chapter 5,
"Request/Response Units (RUs)".
Additional query words: prodsna
Keywords: kbnetwork KB138469