monitor: restarting vtss_appl

[sirhc]

Increasing the switch MTU higher then what is needed poses ZERO RISK. Only a problem if MTU is not large enough, setting MTU smaller than the packet size needed for encapsulation protocols such as VLANs, QinQ, and PPPOE.

Rolling back to v1.4.7 could cause issues, some switches will not support older firmware and configurations could get corrupted.

Personally I think your going down a false rabbit hole but knock yourself out.



Maybe you should read up on PPPOE and how it effects MTU requirements.

I do not use PPPOE but I am pretty sure it increases packet size by 8 bytes so you would need to increase our standard MTU size by 8.

Out standard MTU is 1528 which is enough for VLANs but not PPPOE as well so 1528 + 8 = 1536 MTU

https://en.wikipedia.org/wiki/Point-to- ... r_Ethernet

PPPOE Overhead on Ethernet
On an Ethernet LAN, the overhead for PPP + PPPoE is a fixed 2 + 6 = 8 bytes

[SemperFi]

But I'm not pinging the switch through a PPPOE port. I'm on a directly connected router and I have packet loss.

And that packet loss is directly related to the errors I see in the switch log and it only started after I upgraded.

[sirhc]

But I'm not pinging the switch through a PPPOE port. I'm on a directly connected router and I have packet loss.

And that packet loss is directly related to the errors I see in the switch log and it only started after I upgraded.

Well when you upgrade the switch drops and renegotiates Ethernet connections. You could have a bad cable, bad port, bad something.

Or maybe a simple reboot would cause the issue to go away from a marginal Ethernet connection caused by something like a bad cable/end or noise on the line?

Before I down grade I would at least try a reboot or unplug an plug back in the cable.

[SemperFi]

First thing I did after this started happening was reboot it and then reboot it again, but I'll try another reboot in the early AM to be sure.

[sirhc]

First thing I did after this started happening was reboot it and then reboot it again, but I'll try another reboot in the early AM to be sure.

Well if you did reboots already I would move to checking cables.

As I said in a recent post, twice recently I had to send guys out to towers to simply unplug and re plug in Ethernet connectors as with cold warm cold weather contraction and expansion wiggles things lose.

What pisses tower guys off is having them climb and re plug connectors 200 feet up only to find it was a ground level connection so always check ground connection before sending your guys up especially in COLD weather.

[SemperFi]

The other 6 switches that I originally upgraded to 1.5.0 on Monday but didn't bring forward to 1.5.0RCx are still showing packet loss as well.

There is no PPPOE on that network, it's just VLANs between a few buildings, mostly on fiber, some on copper. Seems unlikely that a bad port/cable is the cause on all of them too.

[sirhc]

The other 6 switches that I originally upgraded to 1.5.0 on Monday but didn't bring forward to 1.5.0RCx are still showing packet loss as well.

There is no PPPOE on that network, it's just VLANs between a few buildings, mostly on fiber, some on copper. Seems unlikely that a bad port/cable is the cause on all of them too.

Agreed (with above)

But we're on v1.5.1 and v1.5.2rcX, I am guessing you have a type O.

You should be using v1.5.1 or v1.5.2rc2

Did you look at the port statistics in the switch by clicking on the little gear to right of the port row and select Port Details from the pop up?

Maybe put up a few screen grabs of the ports in question.

[SemperFi]

Yes a typo. Whatever the latest RC is is what this switch is running.

[SemperFi]

Sorry clicked submit on the post above.

These are from the Fiber/Ethernet network that was upgraded from 1.4.7 to 1.5.0 on Monday 2-3% packet loss consistently. Previously 0% packet loss

[sirhc]

There are no CRC or any errors on the ports to indicate a problem on the links to indicate a reason for packet loss.