We're trying to help a customer of ours having trouble applying this fix on a 2.4GHz EPMP 1000 GPS AP.
When uploading the APPX-firmware-2.6.1-RC1-DDR2-EPTP.tar.gz file we get an error message: "Error while trying to read upgrade status"
We have tried upgrading to 2.5.1 first, 2.6 first, and receive the same error each time.
Please advise!
When can we expect 2.6.1 final? We have several AP's affected by this, and I would rather wait until the final release (which is due this week?) to come out before upgrading.
Also - should I have a problem upgrading AP's and SM's from 2.3.1 and 2.4.1 straight to 2.6.1?
do a reboot before you put the upgrade. What's i see when the ap start to glitch that's affect alot of thing random in the AP before they get a kernel panic and reboot by itself
@Josh Baird wrote:When can we expect 2.6.1 final? We have several AP's affected by this, and I would rather wait until the final release (which is due this week?) to come out before upgrading.
I think you're a bit optimistic - unless I read things wrong, the hope is that the public beta will be available later this week, NOT the final release.
j
So I have been able to install 2.6.1 on several of MY APs many of the issues have been resolved.
The problem I'm still having is APs that won't even begin to boot. The Ethernet goes up and down every few seconds, but never fully boots. I grabbed two new APs, installed 2.6.1, and replaced the two I was having prblems with. SAME THING!! I know it can take up to 20 minutes, but after an hour they never booted.
As a last ditch I unplugged the radios from the CMM and plugged them into the power supply that came with the APs and eventually they booted..
Any thoughts on why the CMM Micro won't boot the radios, but the PSU will??
We did try rebooting them. We were able to finally get the few that were having this issue to take the new firmware after powering them down completely a few times.
Check the voltage output on your CMM,, the Cambium AP datasheet states 23-56 volt, but I think some of the sync injectors run close to that 23 volt, so you may have some voltage loss on your cables depending on length, etc.
Also I'm wondering if when first booting up it's turning the heater inside the AP on and possibly drawing more power than what your injector is supplying.
We have seen this with using Mikrotik powerboxes to power them and eventually they do boot, possibly the heater is done warming the chipset and no longer creating extra power draw?
Good luck!
@dkeltgen wrote:This firmware will run on all ePMP hardware. We're running it on all of our radios.
Dan Sullivan send me a list of compatible MAC address. Not sure if it's the same firmware you had but mine was not compatible with every device.
Hi,
ePMP Beta Release 2.6.1-RC5 is now available. This release addresses the cold issue along with a few other critical bug fixes. This release is compatible with all ePMP radios similar to 2.6.1-RC1 that was posted earlier in this thread.
Thanks,
Sriram
We had several ePMP 1000 5Ghz radios that were constantly rebooting. All had 'CE' in the mac address.
This firmware has solved the problem.
Thanks!
Should an ePMP running v2.5.1.tar.gz be upgraded to v2.6.tar.gz before upgrading to v2.6.1-RC5.tar.gz? Or can I directly upgrade to the RC5?
You should be able to go directly to the RC5, fixed our issues.
2.6.1 does not seem to have fixed our rebooting issues, which does not seem to be related to cold temps.
debug log:
<1>CPU 0 Unable to handle kernel paging request at virtual address 00000000, epc == c0c8311c, ra == c0c83250
<4>Oops[#1]:
<4>Cpu 0
<4>$ 0 : 00000000 00000001 00000000 00000000
<4>$ 4 : 00000000 878d6000 00000001 871002c0
<4>$ 8 : 00000008 00000040 00000000 8710e940
<4>$12 : 00000066 00000003 51eb851f 002625a0
<4>$16 : 878d6000 878d0000 00000000 878d0000
<4>$20 : 878d69bf 00000008 00000008 00000000
<4>$24 : 00000000 8007f590
<4>$28 : 802b4000 802b5b28 00000000 c0c83250
<4>Hi : 00000455
<4>Lo : c28f5ff4
<4>epc : c0c8311c _ieee80211_free_node+0xcc/0x334 [tdd_umac]
<4> Tainted: P
<4>ra : c0c83250 _ieee80211_free_node+0x200/0x334 [tdd_umac]
<4>Status: 1100ff03 KERNEL EXL IE
<4>Cause : 0080000c
<4>BadVA : 00000000
<4>PrId : 0001974c (MIPS 74Kc)
<4>Modules linked in: tdd_umac tdd_link_test(P) tdd_dev(P) tdd_spectral(P) tdd_d fs(P) tdd_hal(P) tdd_acm(P) tdd_unblockSA(P) tdd_pps(P) gpslic tdd_cdf tdd_adf t dd_asf(P) athrs_gmac tdd_netlink_socket(P) macvlan nf_conntrack_netlink ebt_ulog ebt_ip ebt_arp ebt_vlan ebt_pkttype ebt_limit ebt_802_3 ebtable_filter ebtable_ broute ebtables nfnetlink nf_conntrack_tftp xt_HL xt_hl ipt_ECN xt_CLASSIFY xt_t cpmss xt_statistic xt_DSCP xt_dscp xt_quota xt_pkttype xt_physdev xt_owner ipt_R EDIRECT ipt_NETMAP ipt_MASQUERADE iptable_nat nf_nat xt_CONNMARK xt_recent xt_he lper xt_conntrack xt_connmark xt_connbytes xt_NOTRACK iptable_raw xt_state nf_co nntrack_ipv4 nf_defrag_ipv4 nf_conntrack pppoe pppox ipt_REJECT xt_TCPMSS ipt_LO G xt_multiport xt_mac xt_limit iptable_mangle iptable_filter ip_tables xt_tcpudp x_tables ppp_async ppp_generic slhc ts_fsm ts_bm ts_kmp crc_ccitt cambium_iprst leds_gpio button_hotplug gpio_buttons input_polldev input_core [last unloaded: nf_nat_tftp]
<4>Process swapper (pid: 0, threadinfo=802b4000, task=802b6b40, tls=00000000)
<4>Stack : 878d6000 878d6000 871002c0 c0c854f0 802b5b48 878d6000 878d6000 c0c833 d8
<4> 000081f2 00000001 00000030 878d70cc 878d6000 878d6000 878d0000 c0cb59 00
<4> 878d70cc c0cb59ec 87101c3c c0a5efb8 00000000 813b2000 878d70cc 871102 c0
<4> 878d6000 871002c0 8710e50c 00000008 00000002 8710e480 8710e594 c0c53d 90
<4> 00000001 00000000 87101c3c 8710e4f0 0000f37e 00000000 8710e4b8 000000 01
<4> ...
<4>Call Trace:
<4>[<c0c8311c>] _ieee80211_free_node+0xcc/0x334 [tdd_umac]
<4>[<c0cb5900>] tdd_mlme_sta_disassoc_on_ap+0xcc/0x170 [tdd_umac]
<4>[<c0c53d90>] tdd_uq_timeout+0x154/0x34c [tdd_umac]
<4>[<c0c7322c>] tdd_uqt_alloc+0xac/0xc0c [tdd_umac]
<4>[<c0c742e4>] tdd_scheduler_uqt_node+0x3f0/0x604 [tdd_umac]
<4>[<c0c746bc>] tdd_scheduler_uqt_vhl+0x1c4/0x29c [tdd_umac]
<4>[<c0c74984>] tdd_scheduler_ap_ul_qt_schedule_main_queues+0x11c/0x174 [tdd_uma c]
<4>[<c0c74a20>] tdd_scheduler_ap_ul_qt+0x44/0xe4 [tdd_umac]
<4>[<c0c283ec>] ath_tdd_scheduler_start_ap+0xf4c/0x1754 [tdd_umac]
<4>[<c0a397a0>] ath_tdd_sta_ul_stop_timer_handler+0xd0/0x47c [tdd_dev]
<4>
<4>
<4>Code: 8e020014 ac620014 8e020014 <ac430000> ae000010 ae000014 40026000 3 0420001 54400018
<0>Kernel panic - not syncing: Fatal exception in interrupt
Matt,
This crash is not related to cold and it is software induced. We're investigating this crash as our top priority issue. We hope to have a fix for this soon. If your radio is not impacted by the cold issue, you can downgrade to 2.5.2 to avoid this crash.
Thanks,
Sriram
Thanks Jano! Im onsite now. Your answer will have saved me allot of time! I'll update on progress.
RC5 seems to have fixed the problem. All the customers are happy. Time will tell.
All the ePMPs ugraded from v2.5.1 to 2.5.6 RC5 with no problem.
I have also upgraded to 2.6 and I am still having issues. I would like to try 2.6.1 RC5...can i get a link pls.
Thanks
Dan
@Dan DeGirolamo wrote:
I have also upgraded to 2.6 and I am still having issues. I would like to try 2.6.1 RC5...can i get a link pls.
Thanks
Dan
Hi Dan,
Please use 2.6.1-RC10: https://support.cambiumnetworks.com/files/epmp/beta
Thanks,
Sriram
Got it thanks
ePMP Release 2.6.1 open beta is complete and is now available as a full release.