Experimental Physics and
Industrial Control System

"Mark Rivers" <[email protected]> · Wed, 3 Sep 2003 09:12:22 -0500

I just did some more testing.  I added a CA PV link from a "calc" record to another IOC, 164.54.160.94.  The first time I rebooted the crate, all CA links completed OK, there were no "Virtual Circuit Disconnect" messages.  I then simply rebooted again, and observed the following right after iocInit on the client IOC:

DB CA Exception: "Virtual circuit disconnect", context "164.54.160.75:5064"
DB CA Exception: "Virtual circuit disconnect", context "164.54.160.94:5064"
CA.Client.Exception...............................................
    Warning: "Virtual circuit disconnect"
    Context: "164.54.160.95:5064"
    Source File: ../cac.cpp line 1537
    Current Time: WED SEP 03 2003 08:55:25.220365818
....................................................................
CA.Client.Exception...............................................
    Warning: "Virtual circuit disconnect"
    Context: "164.54.160.95:5064"
    Source File: ../cac.cpp line 1537
    Current Time: WED SEP 03 2003 08:55:25.537032472
....................................................................

"164.54.160.75:5064" is a 3.13.7 vxWorks IOC.  The only link to this IOC is an input link of a "transform" record.
INPE:CA_LINK BL13:srCurrent NPP NMS

"164.54.160.94:5064" is a 3.13.7 vxWorks IOC.  The only links to this IOC are 2 input links from a "calc" record.
INPB:CA_LINK 13IDC:m12 NPP NMS          INPC:CA_LINK 13IDC:m22 NPP NMS

"164.54.160.95:5064" is a 3.14.3 vxWorks IOC (upgraded last night from 3.13.7).  It is the IOC to which the seq program is trying to connect 2 PVs.

So the problem is NOT limited to the seq program, all CA links to other vxWorks IOCs exhibit this behavior (although the error messages differ slightly).  Sometimes they connect fine, sometimes they have this initial disconnect right after iocInit.  Once they connect they run forever, I don't recall seeing any flaky operation after the IOCs are running.

I have never seen these disconnects for PVs that are served by our "PV gateway", only for PVs that come from vxWorks IOCs.

The problem appears to be limited to IOC to IOC communication.  All of our Unix/Windows clients connect to these same IOCs with no problems.

Mark

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System