Netonix Forums

I've got a problem I can't explain to myself with a single WS-6-MINI switch and am hoping someone out there has a suggestion...

After replacing the core router that connects this particular switch to the internet (one Mikrotik to another), this switch has started dropping pings and SNMP monitoring of it is dropping as well (matching the ping drops). Lets call this problem switch, "Switch B". It is 2 hops away from the core router we swapped out. One hop away is "Switch A", also a WS-6-MINI - which is not experiencing any issues. There is a point to point wireless link from router to Switch A and another PtP wireless link from Switch A to Switch B. Switch B also has a much slower PtP connection back to the router that is handled via RSTP.

Here is a 30 day Ping graph of Switch B:



The green lines are packet loss. There is also a very regular pattern to the packet loss. Below is the same switch with a 2-day average, not a 30 day average:



Here is the same time frame for Switch A:



No packet loss except for the moment we swapped routers...

Now, where this confuses me more is the pings to the radios attached to Switch B are fine:



That tells me this is not a wireless issue. Nothing shows any packet loss or SNMP issues except Switch B. When I look in the logs, I do not see RSTP flapping during the packet loss periods. The switch is well within temp ratings (that is graphed as well). No firmware was upgraded on the switch. It is running 1.4.9 during this whole period.

There are a bunch of clients hanging off an AP attached to Switch B. None of them show any packet loss or issues. It is only Switch B itself but it started the moment we upgraded our router (no config changes, just a hardware change). I am seeing this behavior being reported to me from two different network management computers (PRTG and Sonar) so it is not a monitoring issue.

Any thoughts?

Try version v1.5.0rc4

We have had a policy of never installing beta or RC firmware in production equipment unless it is fixing a specific problem we are seeing. I'm not seeing anything in the release notes for 1.5.0 that seems to relate. Is this something you have seen before or that you think 1.5.0 will address? We have other WS-6-MINI switches in the exact same configuration as this one (fed from two directions, same number of hops from routers, etc) and they are all acting normal.

When you find an issue we ask users to upgrade to latest code that we are developing tp insure the issue is still there.

v1.4.9 is closed and not on our systems and we do not want to spend time tracking down something that may already be fixed.

Yes v1.5.0 greatly reduced CPU load which handles the CLI/UI and other services like ICMP response from the switch itself and SNMP to name a few.

Please upgrade to v1.5.0rc4 and post up a screen grab of the Device/Status TAB

Also make sure your SNMP server is not polling the switch too often, how often are you polling the switch?

Also do you use our Manager software and of so make sure you do not have more than ONE Manager monitoring any switch or this can happen.

We poll the switch every 30 seconds and ping it at the same time. We use PRTG to monitor.

I'll try 1.5.0 Tuesday when I can gain physical access to it in case it doesn't wake up. I'll report back.

Upgraded to 1.5.0rc4 earlier today. So far, issue has gone away. Not sure if it was the reboot or the firmware but it's been solid for over 6 hours: