My boss and another employee are set to 12 Mbps. Sometimes it runs great, but seemingly randomly their speed will drop to 1 or 2 Mbps. We’ve been fairly careful with the modulation levels that we are willing to accept, and their AP only has about 8 customers. The issues we are having were showing up even when we had 4 or 5 customers on it. We have 3 other APs on this tower and they SEEM to be doing the same thing at the same time. They are all on different 7 MHz channels with at least a 1 MHz guard band between.
I’m wondering if anyone has had problems with load balancing on these. With our 5.7 PMP100 equipment we can have several customers on them, and they still seem to be solid. We just put up another one in Wetumka and so far it has not seemed to have the same problems.
If load balancing is not the issue then I am thinking it is either the CMM4 or the switch. Actually the CMM4 web GUI has never been quite right. I’ve never been able to set up an admin account w/ password. When I go to the Accounts screen, the drop-down list that should contain “admin” contains absolutely nothing. So it’s always had a message that says, “No valid accounts configured. Using default user account” in big red letters near the top of the GUI. I’ve tried resetting it to factory default to none avail.
We have been putting off buying a backup anyway so we will probably be replacing it in the near future to see if that helps, and this time I want to get the rack-mounted CMM, unless someone suggests otherwise. Any recommendations on a good switch?
Are you using the EtherWan/MOXA switch inside the CMM4? What firmware and have you turned off flow control?
Yes we are using the CMM4 w/ the “integrated” EtherWAN switch. 8 x 100 megabit ports for all the CMM ports plus 4 extra, and 2 gigabit.
Firmware for the CMM4 is 3.0. The PMP 320 APs are e2.3. CPEs are f12b00-v8.1.4CAMBIUM.
I don’t know what you mean by that last bit, have I turned off flow control. Does that relate to service flows? Please feed my hungry brain.
EDIT: Switch firmware is: 1.91.0 06/10/10 10:27:48.
Might want to grab the latest firmware for the switch here: http://www.etherwan.com/support-mo.php
What I’m referring to is 802.3X flow control, used to throttle ethernet ports. It causes problems with the PMP320 APs. You need to login to the EtherWan switch and make sure ‘flowcontrol on’ isn’t configured on any of the ports. If I can find the bug I’ll post it here…
Thanks a lot for the help!
Did not update firmware but did disable flow control on all ports on both CMM4s. So far, so good, but the problem wasn’t consistent so I’m still waiting before I proclaim victory.
Good to hear! I think this is where I read about the flow control issue… looks like it’s regarding lockups, but maybe there is a greater effect: viewtopic.php?f=30&t=7053
Would this flow control setting be an issue with a Cisco 3560? My understanding is flow control is disabled by default on Cisco. However I have the same symptoms. 3 people on the tower. 2.3 firmware / synced via ugps / using the default service flows and I get random drops into 1-2 meg range.
Swithport on Cisco shows:
Port Send FlowControl Receive FlowControl RxPause TxPause
admin oper admin oper
--------- -------- -------- -------- -------- ------- -------
Gi0/6 Unsupp. Unsupp. off off 3566206 0
akant wrote: Would this flow control setting be an issue with a Cisco 3560? My understanding is flow control is disabled by default on Cisco. However I have the same symptoms. 3 people on the tower. 2.3 firmware / synced via ugps / using the default service flows and I get random drops into 1-2 meg range.
Swithport on Cisco shows:
Port Send FlowControl Receive FlowControl RxPause TxPause
admin oper admin oper
--------- -------- -------- -------- -------- ------- -------
Gi0/6 Unsupp. Unsupp. off off 3566206 0
Looks disabled by configuration I'm confused as to what's up with all of the RxPause frames. Is Flow Control turned off on other ports as well? (I'm not sure if the RxPause counter will still increment if processing them is disabled)
We are seeing the same issue - “speed auto 10” on our cisco resolves the issue 100%, but then we are of course limited to 10Mbps for the AP. Has anyone else seen this or have any ideas? We have seen the issue on the etherwan switch, 3560-x, 3750-x and 2950 series switches. No errors are showing up on the switches, or the APs for the Ethernet.
I have not seen this issue since I disabled flow control on the EtherWan switch. Speeds are now consistent as long as we keep a handle on our CPE modulation levels.
Nice, good to hear.
That’s interesting regarding 10 mbps.