Created Date: March 26,2019

Created by: Subspace

The following book is for completeness, sorry. I have expanded our internal fiber network at a client site to include two SE-SW8M-2ST managed switches and two SE-SW5U-ST unmanaged switches. I am experiencing communications losses to the system that can be corrected temporarily each time by cycling power to the the affected MANAGED switch. We communicate remotely to the system with AT&T DSL (static IP addressing) with Team Viewer. The system did not have this problem when it was smaller and only had unmanaged switches involved. The setup is as follows: Incoming DSL service connected to Location 1: Managed switch #1 which branches off to two locations: (Location 2) Unmanaged switch #1 with a BRX and EA7 HMI (end of line) and (Location 3) Managed switch #2 with a DL06, EA9 HMI & Laptop computer and one branch to another system (Location 4: Unmanaged switch with a BRX and EA9 HMI (end of line)). When Managed switch #1 locks up (lights appear normal) it kills remote access to everything. When Managed switch #2 locks up I still have access through DSL to Managed switch #1 and the branch to Unmanaged switch #1 . There has been no routine timing of the communications problems and the laptop on Managed switch #2 is used for remote access ONLY (it's in a locked room in a locked building with a crazy password). Nothing else is on this network. The only clue I 've been able to gather occurred with luck when I was onsite one night in the location with Managed switch #2. The lights flickered in the building (Southern California) and I immediately lost communications to Managed switch #2 and everything connected to it. I cycled power to the switch and I was back up and running. Both managed switches have dedicated IDEC 15W power supplies. I have 2 Uninterruptible Power Sources onsite I will be connecting to the Managed switch locations in the near future (in case it is just power flickering issues) but something else seems amiss. Any thoughts? Thank you in advance.

Created Date: March 27,2019

Created by: MikeN

I have no idea why that might be happening but some random ideas:

Could be related to denial of service, either on or off. If its off maybe the switch is getting overwhelmed with packets from PLCs and HMIs. Or if its on maybe it thinks the switch is under a DoS attack from PLC packet spam and blocking traffic.

Maybe a PLC is generating a packet for industrial traffic that is not formed right for normal ethernet use and the managed switches are trying to act on it and locking up when it cant process the packet

Could be some sort of issue when power gets sags during these power blips on the line. Perhaps power is getting low enough to cause hardware lockup but not a full shutdown and when power is back up to normal it still is already locked up

Maybe there is just something not configured correctly in the managed settings and is causing a failure.

Or they could just be sucky switches.