V7.20.5 [stable] is released!

The problem occurs when you want to “push routes”, i.e. you are going beyond allowing simple access to a single remote network from a client. You wan the client to route to multiple networks.

I came across the same situation。It happened

at 7.20/20.1/20.2/20.4,but the CHR run for longer time at 7.20.4.Its kernel failure happened in four days,while other version in one day.

I upgraded a bunch of Mikrotik devices that I manage / support and all fine. I have my home /office network configured with 2 x CRS317 configured in Active / Standby using some proprietary scripts.

The Primary / Active switch gets random packet loss to any downstream devices or ports for few seconds and then comes back. Logs doesn’t show any error nor interface flapping.

When I activate the Standby that has a mirror configuration of the Primary I don’t see this issue at all.

Before this change everything was working fine but after the upgrade I got this issue.

I rolled back to 7.20.4 and the issue remains like shown / detected on the 7.20.5.

I don’t know what else to do more than thinking to a netinstall and/or restore a full backup.

Any tips or ideas?

Thanks!

Yes. Only in CHR.

Maybe you have inadvertently copied the MAC address from active to standby device?

*** SOLVED ***
I found couple different errors that are NOT related with this release. I found an issue with a miss configured interface for my VLANs but also my NAS was kind of stuck with some VMs and until I shutdown those VMs and reboot them the issue went away.

I suspect that there was some sticky MAC Address but I can confirm that ROSv 7.20.5 is working fine now with my setup

********************

Nop, I have scripts that when the standby is active all the interfaces of the primary are disabled. Same on the other direction when the primary is active the standby has all its interfaces down (except the one I use for Heartbeat that is outside of the bridge and has static IPs).

The failover works great, issue is after the upgrade when I active the primary switch I get those random packet loss. Even in some VMs in my proxmox cluster and my physical NAS (that’s the one that gets full packet loss).

I factory reset the swtich and loaded the configuration and the behavior remains. I will plan to do Netinstall today but is totally weird this behavior (and with 7.20.4 and earlier versions from the past year) never had this issue.

Still waiting for the Sandisk USB disks fix, still bugged out. When will it be fixed? I see posts from 2023 talking about it.

You might want to submit a support ticket, with a supout.rif with your particular Sandisk drive attached where isn't not working . While a lot of times some fix for one USB thing may fix others... if it's been two year unlikely it is the "same issue".

>> While a lot of times some fix for one USB thing may fix others...

or else mess it up :slight_smile:

1 Like

They sent me an email yesterday saying that the correction had been made and that a new version would be released soon. See v7.21.rc1 that came out today.

Edit: *) pppoe-server - fixed client disconnects when multiple servers with different service names are active (introduced in v7.20);

I’ve updated Mikrotik RB4011 with 7.20.5 firmware at couple clients and since then SSTP VPN isn’t working properly on those devices, I haven’t changed anything else. I haven’t found solution to it so I used Wireguard for now. Any ideas?

I’m using windows sstp client, it just says it’s connecting forever, says completing connection and then failing. Or just timing out.

sstp_log_mikrotik.txt (17.3 KB)

I saw this release, the problem has been fixed.
Thanks

In version 7.20.5 on IPv6, the upload bandwidth is not restricted.

Anyone else have lots of excessive broadcasts/multicasts, probably a loop since 7.20.x updates?I noticed bunch of errors from various ports and thats on multiple locations and switches/routers and i dont have actual loops..

I get this error also on home HAP AX3 for port ether2 every time i turn on my PC which is connected to it.

Here! Every time a computer is turned off… since v7.19, combined with “possible SYN flooding on tcp port…” so annoying!

1 Like

What is after “port”? “53”?

Version 7.20.6 has been released: