Having HMI to Oracle DB comm problems. Off and on throughout the day 2 of our HMIs will show comm errors. They start out infrequent and continue in numbers until total comm loss.
I am a network admin, and in using the following tools – have not found network problems:
All network switches/routers log to a log server – never a dropped line or protocol
during an event. No Spanning Tree events, never a single event noted in the network logs during an event. No errors on client or backbone interfaces.
Whatsup Gold ICMP pings to all network components, PLCs, and HMIs. During
comm errors – no missed polls to the 70 components. 2.5 second polling interval.
Solarwinds monitors and graphs lower layer errors, throughput, bandwidth usage,
processor percentages – no problems seen.
Sniffer distributed indicates low packet to packet deltas, below 1-5 ms between
packets. During event, deltas rise to 60-200 seconds – the server is always the
slow responder. Specifically on TCP port 1525 for the DB. As such, the team is
focused on this being a DB issue.
Network is all Cisco:
Network Core layer is 2 Cisco 4506 with SUP IV engines
Plant is multiple VLANS – components involved in comm problem are in the same
VLAN.
Network Distribution layer is 2 Cisco 2950C – all meshed fiber connections to
core and access layers.
Network Access layer is 6 Cisco 2955 in NEMA boxes on the floor
Server is: Dell 2650 2 processors at 3Ghz, 2G memory and is running Intellibatch,
DASABCIP & Oracle DB. Meshed fiber to distribution layer, cat 5 to PLCs,
HMIs, etc. Cat 5 cables are all in spec
2 Control logics: 5555
50 PLCs are: SLC505
PLC network modules are: 1756-ENBT/A
2 HMIs are Dolch rugged Win2K, running Intouch 8.0, DASABTCP, Oracle client
Has anyone encountered HMI to Oracle DB comm problems such as I have
described?
Thank you,
Phil
I am a network admin, and in using the following tools – have not found network problems:
All network switches/routers log to a log server – never a dropped line or protocol
during an event. No Spanning Tree events, never a single event noted in the network logs during an event. No errors on client or backbone interfaces.
Whatsup Gold ICMP pings to all network components, PLCs, and HMIs. During
comm errors – no missed polls to the 70 components. 2.5 second polling interval.
Solarwinds monitors and graphs lower layer errors, throughput, bandwidth usage,
processor percentages – no problems seen.
Sniffer distributed indicates low packet to packet deltas, below 1-5 ms between
packets. During event, deltas rise to 60-200 seconds – the server is always the
slow responder. Specifically on TCP port 1525 for the DB. As such, the team is
focused on this being a DB issue.
Network is all Cisco:
Network Core layer is 2 Cisco 4506 with SUP IV engines
Plant is multiple VLANS – components involved in comm problem are in the same
VLAN.
Network Distribution layer is 2 Cisco 2950C – all meshed fiber connections to
core and access layers.
Network Access layer is 6 Cisco 2955 in NEMA boxes on the floor
Server is: Dell 2650 2 processors at 3Ghz, 2G memory and is running Intellibatch,
DASABCIP & Oracle DB. Meshed fiber to distribution layer, cat 5 to PLCs,
HMIs, etc. Cat 5 cables are all in spec
2 Control logics: 5555
50 PLCs are: SLC505
PLC network modules are: 1756-ENBT/A
2 HMIs are Dolch rugged Win2K, running Intouch 8.0, DASABTCP, Oracle client
Has anyone encountered HMI to Oracle DB comm problems such as I have
described?
Thank you,
Phil