How many entries did you have in the list. I had a similar problem but mine always came after restart. My entries where static but the disk on the unit was full and I could add more via script but after a restart I lost them. Never managed to solve this so now I have a startup script that populates everything have have gone for dynamic entries instead.
You seem to be correct, although it doesn’t completely explain why only ~200 entries remained out of 50k (all are/were static). I have many small address lists and a huge one with over 50k entries. There is only 2.1MB free space on hAP ac^2 nand chip, and these are the sizes of a couple of latest backup files:
-rw-r----- 1 che che 203K Aug 1 04:00 2020-08-01_hap_ac2.backup
-rw------- 1 che che 2.9M Aug 1 04:00 2020-08-01_hap_ac2.rsc
-rw-r----- 1 che che 224K Sep 1 04:00 2020-09-01_hap_ac2.backup
-rw------- 1 che che 2.9M Sep 1 04:00 2020-09-01_hap_ac2.rsc
I will just remove the big acl, since I do not want to test if it will wipe the configuration after restart (router was not restarted in months until today update).
Can I downgrade to this from 6.47 without losing the config?
Yes, make a backup to feel more comfortable anyways
The problem is not the backups (they are done automatically every day), but that these devices are in a remote location, so I can’t access them physically.
There’s an bug with LTE on startup on CCR1016. Static ipv4 address(es) are loosing their interface setting in IP addresses and coming up ‘unknown’.
This causes routes to break to the LTE modem as there is no address on the lte1 interface without correcting the unknown and putting back lte1.
Happened twice. Once on upgrade to 6.46.7 and again to reboot to update the bootloader so its completely reproducible. Means backup LTE is pretty useless without access to IP addresses.
Ticket SUP-27850
I have a problem discovering devices. In a small network, the main router ccr 1016, discovery interface is bridge1,
it is also configured as OVPN server, type Ethernet, OVPN profile - bridge1. There is a remote OVPN client - rb951,
its discovery interface is also bridge1. So, in ip neighbors, 1016 sees the entire network except rb951,
and rb951 also sees in ip neighbors the entire network, except for 1016.
in 6.45.9 there was no such problem
Yesterday I caught a joint at 6.46.7 !!!
Ping breaks from the dhcp server to the client, and from the dhcp client to the dhcp server, ping is lossless. At the same time, when connecting to an access point, the Internet is there for a couple of minutes and then it will be irretrievably lost !!!
Returned 6.45.9 again everything worked in normal mode.
Equipment:
Hap ac lite - dhcp server.
CAP ac - dhcp client.
Had to roll-back to 6.45.9 on my Rb4011. Have had issues which only we’re solvable by rebooting the router. Couldn’t make supout.rif file either before reboot. The last problem was all non local routes were gone, couldn’t see them in winbox or terminal. Actually the route list was totally empty but locally connected routes worked. No OSPF or BGP connected routes worked. Before that I’ve had one random reboot of the router. Before the upgrade it was totally stable and been running for like 100 days uptime or so.I’ll let the rest of the equipment run on 6.46.7 (hAP, cAP, cAP AC, wAP ac, CRS309, CRS317) since they seem to work fine. So I’ll stick with 6.45.9 on the RB4011 for a while now. Anyone else seen any strange behavour with this version?
I upgraded a new 328-24P-4S+ to this release. When setting a bridge VLAN in Winbox the new window shows a VLAN ID of 4294967295 and ‘VLAN IDs’ is red (since number is invalid). I change it to the VLAN ID I want and it seems to be fine. It shows this each time I add a new VLAN.
I confirmed this happens on a second 328, and on a 326. Both upgraded from previous long-term 6.45.9.
It looks like this VLAN entry thing is happening on other devices as well - even when configured from scratch. Configuration seems to work OK when it is done.
This sounds very scary. Have you contacted support as you seam to be able to replicate the issue?
Seems like i will stuck at 6.45.9, because this one have some issues. Wait for another “stable” long term build.
Hi,
This is my post in winbox release thread: https://forum.mikrotik.com/posting.php?mode=edit&f=21&p=816318
Straight about “red vlan error”
Yes, I had to downgrade to 6.46.6, w60g links were very unstable with 6.46.7, especially Wireless Wire, but some LHG links too.
I can confirm this issue. Upgraded some of our LHG60’s to 6.46.7 and we’ve had multiple disconnects since. Downgraded them to 6.45.9 again and they are rock solid.
I can confirm this. I had to revert my Wireless Wire units back to 6.46.6
in DHCP-client scripts don’t work
In our CCR1036 devices, DHCP does not always populate the ARP table. ARP table entries are critical in our setup.
Or the ARP table entries are disappearing despite the longer timeout setting on the DHCP bridge.
Actually, thinking more and more that the ARP entries get added to the ARP List but later disappear for some reason. Despite the long timeout setting on the Bridge.
in CRS305-1G-4S+ (ARM) v6.46.7
SFP+DAC of any brand not work in 10G mode
UPDATE:
not work between Huawei MA5608T into 10GE port and CRS305-1G-4S+, with mikrotik dac, Ubiquiti Dac, works only with some models of optical sfp+ transceiver
(I deduce that it may be some driver missing from his ARM platform)
works well in anothers CRSxx models witdh dacs and sfp+