A TCP/IP communications port controlled by SOLVE:CPT presented slowness. Cust checked with
IBM on TCP/IP but the cause was not identified. It was shown delay on some moments
of connectivity controlled by SOLVE:CPT.
Also a gap (drop) was noticed on the connections using the SOLVE:CPT LISTEN port. Following
each gap (drop)--that occurred every 4 hours--work resumed normally.
Following the MIPS addition to the environment and creation of a new CICS
the problem went away.
On the gap (drop): It was discovered that the problem was on the server
that calls CICS. Every 4 hours the switch 're-learns' the ARP.
Situation was corrected.