I have found what might be a bug in the TR-069 client. If I try setting any parameter to a value with an ampersand (ex. changing identity from MikroTik to Mikro&Tik) through my ACS (GenieACS) it goes crazy and gets stuck in an endless repeating communication loop trying to set the value over and over again unsuccessfully hundreds of times per second until I not only cancel the task in GenieACS but also disable and re-enable the TR-069 on the MikroTik.
The GenieACS developer was not able to reproduce this problem with his non-MikroTik devices. This makes me wonder if it is a bug in the TR-069 stack on MikroTik - can any TR-069 users reproduce this issue?
That’s additional data to be sent over the wire, which may be significant on a busy system with lots of registered clients. I believe keeping the SNMP traffic as low as possible by eliminating the duplicates is a good thing.
Update: The GenieACS developer says there is a bug in MikroTik’s TR-069 client. Any SetParameterValues request that includes an ampersand in the value causes an empty reply to be sent to the ACS by the MikroTik. This in turn causes a problem in GenieACS which is not expecting an empty reply from the MikroTik and resends the request, causing a loop.
Changes since previous version:
*) gps - added “fix-quality” and “horizontal-dilution” parameters;
*) snmp - fixed CAPsMAN registration table OID print;
*) snmp - fixed CAPsMAN registration table SSID type;
*) snmp - optimized bridge table processing;
*) tr069-client - added “Ethernet.Interface.{i}.MACAddress” parameter;
*) userman - allow access to User Manager users page only through “/user” URL;
If you experience version related issues, then please send supout file from your router to support@mikrotik.com. File must be generated while router is not working as suspected or after crash.
I saw this on RB750Gr3. System / Routerboard / Settings. It’s possible to see and change RAM frequency via CLI but not via Winbox.
CLI keeps telling me there is a RAM overclock when I run /export, now running at 1200MHz and I didn’t changed it.
There is no change log available for Routerboot version 3.35.
Running 6.38.1, posting here because I think this post it’s a good place to talk about bugs. If not, I’m sorry.
I am asking myself, whether it is possible to get the values of the amount of users, connected to a specific SSIDfor a single CAP with SNMP polling the CAPSMAN controller?
Looks like you guys got it this time! Thanks for the quick fix.
Still sad you are being stubborn about leaving/adding back the mac OID in there. I think redundancy is a poor argument and believe myself and others will benefit from this being available. So many standard MIBs, including ones you are using (like bridge mib) provide this even if the mac is encoded in the OID in decimal format. Also, using the bridge and arp tables as previously mentioned in this thread are incomplete workarounds in my opinion since the entries in those tables can timeout (or never exist depending on configuration) even when a wireless client is still connected. Most tools don’t have a built in way for parsing and converting the mac address from the OID, rather they rely on SNMP index and data provided in the table.
IMHO the MAC should always be encoded in the OID as 6 decimal numbers, not as a binary string.
It causes difficulty in scripts that are not binary transparent.
I think it was first started by copying some OID values used by Ubiquiti e.g. for reporting link parameters on an AP by client.
But using 6 decimal numbers instead of the 6-byte binary is much better.
Does the black list function in the discover info window of the dude client for windows work right in this build? In 6.38.3, and previous versions I don’t see the . or … buttons shown in the manual here.
Changes since previous version:
!) routeros - added support for new products and RouterOS features which will be announced at MUM EU (https://mum.mikrotik.com/2017/EU/info/EN);
*) capsman - fixed “/caps-man manager interface” compact export;
If you experience version related issues, then please send supout file from your router to support@mikrotik.com. File must be generated while router is not working as suspected or after crash.
This version (Edit: no necessarily this specific version, but rc4 is perfectly fine for example) completely broke DFS for me in Germany. My wAP AC resides on the top floor, sometimes no 5GHz clients connected. As soon as I come near it with a client (tested with iPhone 6s, multiple times) it logs something like pasted below. I’m pretty sure it interprets clients looking for WLAN APs as RADAR. Please fix this! It’s the same for a hAP AC with the current stable version (v6.38.3).
Feb 26 10:06:12 mikrotik-wap-ac wireless,info wlan2: radar detected on 5580000
Feb 26 10:06:17 mikrotik-wap-ac wireless,debug wlan2: must select channel
Feb 26 10:06:17 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5580000
Feb 26 10:06:17 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5260000
Feb 26 10:06:17 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5580000
Feb 26 10:06:17 mikrotik-wap-ac wireless,debug wlan2: selected channel 5500000
Feb 26 10:06:17 mikrotik-wap-ac wireless,debug wlan2: search for radars on 5500000
Feb 26 10:07:17 mikrotik-wap-ac wireless,debug wlan2: no radar detetected, start network
at 10:07 I was already back downstairs, going up again and leaving the phone there the following happens:
Feb 26 10:20:14 mikrotik-wap-ac wireless,info wlan2: radar detected on 5500000
Feb 26 10:20:19 mikrotik-wap-ac wireless,debug wlan2: must select channel
Feb 26 10:20:19 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5580000
Feb 26 10:20:19 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5260000
Feb 26 10:20:19 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5500000
Feb 26 10:20:19 mikrotik-wap-ac wireless,debug wlan2: radar reported on 5580000
Feb 26 10:20:19 mikrotik-wap-ac wireless,debug wlan2: selected channel 5180000
after selecting a non-DFS channel I see my iPhone connecting to the AP in the logs.
Now that TR-069 support is basically here and working, I would like to request a change to the RouterOS Main Package:
Could you please have the IPv6 package enabled by default instead of disabled? Or at least provide a means of enabling the package in NetInstall?
We want our TR-069 managed customer routers to have IPv6 enabled and configured when they receive them - the way things are set up currently, the customer router needs to be rebooted after they boot it up for the first time in order to enable IPv6 and apply the IPv6 config. Our defaults script applied by NetInstall can enable the IPv6 package but can’t configure the settings until the router is rebooted because the package isn’t active yet. This makes things much more awkward for us.
alexjhart, MAC-address is not present in /interface wireless registration-table print oid either, and it is calculated in the same way as for capsman table OIDs.
