SG500x Point to Point Connection very slow after 1.30.59 update

[Mike Horowitz]

Hi,

I have a point to point 100mb line connecting SG500x switches at 2 locations, with these ports in a vlan together. I use layer 3 on the SG500x's to route between the two sites. I also have RIP protocol enabled to allow for automatic failover if the point to point fails (this works well).

I upgraded the ios to 1.30.59, and the connection now runs at a fraction of the speed. I have hard-coded the ports to run at 100mb full duplex, but I still only get around 12mb of throughput between the switches. I used to get about 90mb before. The provider also hard-coded the circuit to 100mb full duplex. The provider has checked the circuit remotely, and everything looks fine. I tested with a laptop at each side, and did get almost 70mb throughput. I would think this indicates some type of setting mismatch, but i cannot find anything.

I am using a backup vpn for now as some things look flaky as well (on vpn, I can use the web interface to access the SG500x on the opposite side, but on the point to point I cannot - I can ping the device though).

I have seen some references on here to shutting off STP on these 2 ports. Not sure why or if that would help.

Any ideas?? Thanks in advance!

Mike

EDIT: Looks like it may be the MDIX setting. Will report back tomorrow if it works...

Message was edited by: Mike Horowitz

[Tom Watts]

Hi Mike, as you might have figured out, setting speed and duplex to a manual setting would be a good first step assuming the provider will do the same on their parts.

This may also sound flaky and I imagine you've heard it before...I'd always recommend to factory reset the unit after a firmware upgrade. Too many times I've seen things like code overlap and other anomalies that simply go away with that precaution. I'd recommend to back up your config, reset the switches and see what happens.

-Tom
Please mark answered for helpful posts

[Mike Horowitz]

Hi Tom,

I had done a factory reset a few months back. I really, really don't want to go down that path if I don't have to :-( . I did restart the switches. The connection does work, which is wierd (but at the much lower speed). I am going to have the line provider come on-site to test their equipment further just in case.

The MDIX hard coding did not work. I am not sure what else I can do on the switch. What I do know is that the connection stopped after I upgraded the first switch at location #1, but reconnected at the slower speed after the second location was upgraded.

Thank you,

Mike

[Mike Horowitz]

Problem solved. Something happened to my failover RIP tables during the upgrade, and had to be rebuilt. All is good now. Data was being passed over vpn, and also point to point it appeared. Thanks again Tom!