10-08-2018 05:39 PM
i hav a AF5XHD link that is 100% stable with Site A as master and Site B as Slave...
I Flip Site B to Master and link fails in short order (5-30min)
Site B Master has the following error in the Log
Oct 8 19:55:53 air_fpga: [mac/3] ap_hb_resp_timeout(1973) hb timeout, cpeId:0, retry:4 Oct 8 19:55:53 air_fpga: [mac/5] ap_hb_resp_timeout(1979) [DISASSOC]: Ready to disassociate the CPE 0 Oct 8 19:55:53 air_fpga: [mac/5] ap_disassociate_cpe(533) AP is going to disassociate CPE id 0, cause 6 (DISASSOC_DUE_TO_HB_TIMEOUT or SCAN) Oct 8 19:55:53 air_fpga: [mac/5] ap_free_cpe_resource(411) Free resource for CPE id 0 Oct 8 19:56:37 air_fpga: [mac/5] macsm_mute_dfs(54) [ap_usr_disassocing_handlr:1390] MUTE DFS (cpe 0)
Slave Gives error in log
Oct 8 19:56:24 air_fpga: [mac/3] cpe_hb_wait(2000) Where is my hb? state(6) Oct 8 19:56:24 air_fpga: [mac/5] cpe_disassociate_ap(955) CPE is going to disassociate connected AP, cause 6 (DISASSOC_DUE_TO_HB_TIMEOUT or SCAN) Oct 8 19:56:24 air_fpga: [mac/5] cpe_disassociate_ap(966) <send> DISASSOCIATION msg Oct 8 19:56:24 air_fpga: [mac/5] cpe_free_resource(752) cpe clean resource done Oct 8 19:56:24 air_fpga: [mac/5] cpe_hb_wait(2011) Do Connect Freq (5745)
Using GPS sync
Does not matter if 20Mhz or 40Mhz spacing
Tried Two different AF5XHD at site B with same results link....
Second AF5XHD at Site B is running a stable link as Master without any issues to Site C 3km away which is located between Site A and B (7.5km link)
Confirmed that no other radio on either tower A, B or C are operating near the frequency of the link (Verified radio frequency and on AirView)
All radios runing 1.1.1 Firmware
Ubnt Tech support has not given me any help in 7 days except ask me to take down this Main Production link and swap the radios on the 100ft Towers and see if that fixes the issue.
I need GPS syn on Tower B as its my main distribution Tower.
10-09-2018 01:45 PM
Your expected power is -38 while your actual signal is -68.
When the airLink Expected Signal and the actual signal do not match, you might see any sort of issues that can be very difficult to solve without getting a better signal.
Once you have a matching expected signal and actual signal, please try swapping the master and slave again and let me know how it goes.