v1.3.9rc2 through rc14 - BETA/Release Candidate

[lligetfa]

Installed rc6 on WS-6 and WS-24.

[tma]

I can confirm that both the "Broadcast storm when using LACP and RSTP" and "Unable to change RSTP path cost for LAGs" problems are solved in 1.3.9rc6.

[mayheart]

Running v1.3.9rc6 on a few switches, no longer having SNMP problems.

[designweb]

installed on 15 various switches all looking good, continued thanks

[Dawizman]

Well, SNMP went down tonight on the switches that I haven't yet upgraded to 1.3.9, but the one I did is still working properly. I guess that means the SNMP bug is officially gone!

Waiting on a stable release now to upgrade the remainder of our switches.

[Dr_Agony]

So the release notes say that "hundreds and hundreds" of MAC addresses aren't an issue any more, but I typically run closer to 3000-5000 MAC addresses. I haven't deployed this switch yet, because without a management VLAN it's useless to me. But now I'm wondering if I should even try... will it "fall over dead" with that many MACs?

[sirhc]

So the release notes say that "hundreds and hundreds" of MAC addresses aren't an issue any more, but I typically run closer to 3000-5000 MAC addresses. I haven't deployed this switch yet, because without a management VLAN it's useless to me. But now I'm wondering if I should even try... will it "fall over dead" with that many MACs?

It does have a management VLAN, not sure why you say it does not?

You can specify the management VLAN ID, the default is 1 but you can change it to whatever you want it to be.

The management VLAN is the top VLAN, you can not delete it but you can give it any ID or description you want.

I do not know how it will react with 5,000 MACs as this is pretty hard for us to test.

I think Fransisco had a couple thousand and it worked for him with the v1.3.9rcX.

It is not like it crashes it just is slow to report the MAC table. This is because we provide additional information that most switches do not like the manufacturer and last known IP address which has never been done before on a layer 2 switch.

However if you find it does not work let us know by reporting back here and maybe providing us access to the unit via TeamView to see what we can do.

[sirhc]

Personally I can not imagine running one LARGE flat network (5,000 devices on one flat segment) like this, must be a nightmare if something goes wrong like a packet storm?

[sirhc]

We finally fixed UBNT Discovery, sorry for the delay.

We are on schedule to release v1.3.9 FINAL maybe by Sunday or Monday

Things being done yet for v1.3.9 FINAL:
Ability to disable Safe Mode, requested by Adair

Input Fields better protected against invalid entries

Log file formatting will allow you to cut and paste and retain proper formatting

SAFARI check for compatibility

UI Enhancement for MAC table so it tells you the number of MACs in the table

Add clear stats command to CLI

IPERF, determine if we will keep it and further develop it. The CPU in the switch is not real fast and will never be able to generate 1G of performance but if we determine that it can not even produce 100M of traffic we may ditch it. REMEMBER THE CPU IS JUST THERE TO DRIVE THE UI/CLI NOT SWITCH PACKETS.

Incompatible SFP Module request by user, a 3COM SFP module - May get moved to later version

[lligetfa]

RC7 installed on WS-6 and WS-24 with no issues. Discover works. Linux log fixed.