In an environment with multiple network cards, the Slump binds to the wrong network card. We expect the Slump to use the first of the bind list.
One of the symptoms is that the Service takes too long to start. Sometimes it fails, sometimes the service starts.
However, even after starting, the SDM UI is not accessible and in the stdlog file we can see several error messages.
An example of the stdlog entry is:
SIGNIFICANT boplgin.c 463 Can't connect to server 'domsrvr'; will retry in 30 seconds
SIGNIFICANT daemon_obj.c 3268 Waited 120 seconds for domsrvr to connect to slump.
SIGNIFICANT bopauth.c 105 Can't connect to server 'domsrvr'; will retry in 30 seconds
SIGNIFICANT api.c 1916 Fast channel failed
ERROR server.c 3720 Unable to handle fast-channel request from 5762|prov#3088_bpvirtdb_srvr to 5762|bpvirtdb_srvr; reason: Process bpvirtdb_srvr is not currently running
ERROR daemon_obj.c 1990 Daemon pdm_intrvlog_nxd died: restarting
This is caused by a Problem in Service Desk Manager where it will attempt to use both network cards.
Service Desk 12.9 / 14.1 / 17.1
There is a patch available to solve this related to Problem USRD#2726. Please contact CA support to obtain this fix.
NOTE: This patch is already included in the 14.1 CP4 patch update and is part of the baseline code of release 17.0 and higher
WIN-INCORRECT NETWORK INTERFACE PICKED UP
On a server with multiple network interfaces, it is possible for Service Desk Manager to pick the secondary network interface while initializing fastchannel connection.
Enabling logging on tcp_port.c would log messages similar to:
"tcp_port.c 546 Target ip is XX.XX.XX.XX, and addrInfo->ai_family is 2" where XX.XX.XX.XX would be the IP address of the secondary network interface.
As a workaround, to solve the error, follow these steps: