SF-300-48P POE is power cycling

[david.baumgarten]

Hello,

I have a SF-300-48P that is power cycling all POE ports (in use) exactly every 30 minutes. The other older cisco POE switches we were replacing did not have this issue. We have about 40 cisco SPA942 IP phones connected to it. Power consumption is about 175 watts total.

This is my second switch, previously tried a SG500X-48P with the same result.

Any ideas how to stop it from cycling?

Thanks for the help,

David

[Tom Watts]

David, can you confirm these phone models? This may be critical information.

-Tom
Please rate helpful posts

[david.baumgarten]

Port 1 is a SPA504G

Port 2 is a SPA942

I also found an SPA962 but it is on another switch.

[david.baumgarten]

To recap my evening, Port 1 ended up going down. I tried isolating  the switch and just having phones connected to it (not being able to  register). They seem to last about 55 mins before the ports are cycled.

Currently, I have 1 phone plugged into the switch and will check it in the morning to see if it cycled as well.

[david.baumgarten]

Morning update,

The single phone (SPA942) I had plugged in did not reboot at all last night. This morning, I have connected the switches back together and moved 1 additional phone to the new switch.

When I connected the other switch I got an error stating "Duplex Mismatch on FE24". I changed this port to 100mb Full (flow conrol is off).

[david.baumgarten]

No matter what I change, I cannot get ride of:

%CDP-W-DUPLEX_MISMATCH: Duplex mismatch detected on interface fa24., aggregated (1)       

[david.baumgarten]

Another unusual discover is this morning 1/2 of the phones showed to have power but not network. Screen on the phone was blank but buttons lite up. Looked at the switch and it showed to be giving POE but not network signal.

This happened on both switches.

*EDIT*

This happened again and this time I noticed the power consumption for the phones that usually draw 5000ish mw was down to about 3300mw. This was an idicator that the phones were dead but not dead enough to reboot.

[Tom Watts]

Good morning David, please move the switch to 1.2.7.76 firmware then factory reset the switch. Also please make sure the Spa5xx phone is using 7.5.3 firmware.

Can you also move the switches location to a different power source and different surge protector/ups?

-Tom
Please rate helpful posts

[david.baumgarten]

I moved the switches to a new power sources and should see if it worked in about 15 minutes. Then I will upgrade firmwares.

Also - the old switch has a custom VLAN on it. Its a basic lan configuration. Would this conflict with the out of the box vlan settings on the new switch?

BTW, the old switch is a srw224p.

[david.baumgarten]

Switch firmware upgraded, phone appears to already be on the latest firmware.

Different power source resulted in same outcome. I have disconnected the new switch from the other to see if the phones reset or just go dead.

Also - I plugged in a POE access point and it shows a stead consumption of 7300mw. This stays constant even when the phones drop off.

Any other ideas?

[Tom Watts]

The SPA5xx phones should have cdp min/max benefits for the POE, I don't think the old SPA9xx support cdp (only LLDP). The vlan shouldn't make a difference. But it may conflict with smart port. If you have the auto voice vlan set up the switch macros will set the port as needed (data vlan untag, voice vlan tagged) for the at least the SPA5xx phones.

Suffice to say, this is really odd, the reason being, using SPA9xx or SPA5xx phones, I can't produce this symptom using 1 phone or 8 phones. The SPA phones are a SB product line and interop extremely well to the SX300 series.

I can't help but to think there is an external factor since we even have UC deployments using upwards over 100 phones and the same equipment (spa5xx).

-Tom
Please rate helpful posts

[Tom Watts]

David, I have received report this has been seen before.

This is what must be collected;

Topology -> Network diagram

Config file of the switch

# of phones and models

show tech

Firmware on 1.2.7.76

Type of ethernet used

Additionally, if you could try changing the Priority for the POE, by default it is LOW, try to set to the high and critical to see if it stabilizes at all.

If you could, please send me an email with all of this information including your Cisco ID and serial number of the switch (or switches) if you have both the 300 and 500.

Additionally, I am posting a case number here for a different instance where the customer has 40+ Alcatel phones on a SX500 which has similar behavior you've reported.  624095227. This is basically for tracking.

-Tom
Please rate helpful posts

[Bar1s]

Hi David,

Did you solve your case ?

I have something similar problem.

Regards,

Baris.

[Raul Ricano]

I am seeing this issues across multiple small business switches> SF-300-48P, ESW520 and SG500-28P.  Thought it was related to Green Ethernet, but after turning this off it seems to have helped some, but not completly solve the issue.  Did you guys ever come up with a solution to this?