v1.3.9rc2 through rc14 - BETA/Release Candidate

[sirhc]

v1.3.9rc13 - Released 1/7/2016
- fix for clearing log
- added "clear log" CLI command

[Eric Stern]

Hi Eric,
first of all happy new year.
Regarding the issue it's quite simple, add the switch to dude with the correct snmp community, double-click the device you just added, go to snmp->interface and as soon you go there the cpu goes to 100% and snmp become unresponsive, same thing if you make a link (right click-new link), as soon as you select the interface to graph, cpu 100%. Snmpwalk is ok.
Actually i'm having this problem on all WS-12-250A, strange thing is the WS-12-250-DC are working fine (same fw of course).

[/quote]

Hmmm, I still can't duplicate this. I've tried 1.3.9rc6, 1.3.9rc8 and 1.3.9rc11 on a WS-12-250A. I've tried both v1 and v2c for SNMP.

I know that several other people have used The Dude and no one else has mentioned a problem. What version of the The Dude are you using?

[sirhc]

We would like to finalize v1.3.9 FINAL so if some of you can please beat up v1.3.9rc13 and let us know we would greatly appreciate it.

We are excited to start releasing v1.4.0rcX that has some EXCITING new features, here are a couple:

1) QOS

2) Discovery Tab, there is a discovery Tab where you can have the switch look for other devices using UDP, CDP, LLDP.

[lligetfa]

v1.3.9rc13 - Released 1/7/2016
- fix for clearing log
- added "clear log" CLI command

Installed rc13 on WS-6 and WS-24 with no issues. Log clearing worked.

[zaltec]

Hi Eric, Dude is 4.0b3.

Hmmm, I still can't duplicate this. I've tried 1.3.9rc6, 1.3.9rc8 and 1.3.9rc11 on a WS-12-250A. I've tried both v1 and v2c for SNMP.

I know that several other people have used The Dude and no one else has mentioned a problem. What version of the The Dude are you using?

[zaltec]

I just installed a new DC switch with rc13 and it has the same problem, just as soon as you add a new link in dude this is the result:

top - 09:36:37 up 1 day, 18:15, 0 users, load average: 1.07, 0.68, 0.47
Tasks: 56 total, 3 running, 53 sleeping, 0 stopped, 0 zombie
Cpu(s): 92.7%us, 5.5%sy, 0.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 1.8%si, 0.0%st
Mem: 126348k total, 48660k used, 77688k free, 4364k buffers
Swap: 0k total, 0k used, 0k free, 17196k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1093 admin 20 0 5468 1692 864 R 84.5 1.3 0:58.18 snmpd
707 admin 20 0 2112 628 452 R 6.9 0.5 0:01.26 top
1360 admin 20 0 7016 2048 1056 S 3.4 1.6 0:58.05 lighttpd
2658 admin 20 0 19376 13m 1044 S 1.7 10.8 22:18.66 status_thread


On a side note, when STP is disabled how BPDU packets are handled?

[Eric Stern]

I just installed a new DC switch with rc13 and it has the same problem, just as soon as you add a new link in dude this is the result:

top - 09:36:37 up 1 day, 18:15, 0 users, load average: 1.07, 0.68, 0.47
Tasks: 56 total, 3 running, 53 sleeping, 0 stopped, 0 zombie
Cpu(s): 92.7%us, 5.5%sy, 0.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 1.8%si, 0.0%st
Mem: 126348k total, 48660k used, 77688k free, 4364k buffers
Swap: 0k total, 0k used, 0k free, 17196k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1093 admin 20 0 5468 1692 864 R 84.5 1.3 0:58.18 snmpd
707 admin 20 0 2112 628 452 R 6.9 0.5 0:01.26 top
1360 admin 20 0 7016 2048 1056 S 3.4 1.6 0:58.05 lighttpd
2658 admin 20 0 19376 13m 1044 S 1.7 10.8 22:18.66 status_thread

In a previous message you said you are making changes to snmpd.conf. Are you still doing that?

On a side note, when STP is disabled how BPDU packets are handled?

They are dropped.

[zaltec]

In a previous message you said you are making changes to snmpd.conf. Are you still doing that?

Nope, the changes i did were to find where the problem is, but i HAVE to graph ports, right now i'm trading monitoring features with power switching features, i wouldn't call that a bargain.

They are dropped.

That's a VERY big problem, is there a way to make the switch true transparent as it should?

[sirhc]

They are dropped.

That's a VERY big problem, is there a way to make the switch true transparent as it should?

I will let Eric comment on this one but as I understand it only "dumb" non manageable switches will pass BPDU packets?

[sakita]

Typically what I've seen is that BPDUs get dropped when spanning tree is disabled... whether on a switch or on a port (e.g. disabling spanning tree on a port so that BPDUs from an attached radio that doesn't allow spanning tree to be turned off or even adjusted - kills the problems caused by wrong spanning tree values).

In short, I would consider dropping the BPDUs normal behavior.