v7.1.4 and v7.1.5 is released!

emils · Tue Mar 22, 2022 10:20 am

RouterOS versions 7.1.4 and 7.1.5 has been released "v7 stable" channel!

Before an upgrade:
1) Remember to make backup/export files before an upgrade and save them on another storage device;
2) Make sure the device will not lose power during upgrade process;
3) Device has enough free storage space for all RouterOS packages to be downloaded.

What's new in 7.1.5 (2022-Mar-22 13:03):

*) route - fixed "table" menu emptying after RouterOS upgrade;

What's new in 7.1.4 (2022-Mar-21 13:23):

*) bgp - fixed VPNv4 route sending to remote peer;
*) bridge - fixed destination NAT when using "use-ip-firewall" setting;
*) bridge - fixed filter rules when using interface lists;
*) bridge - fixed priority tagged frame forwarding when using "frame-types=admit-only-untagged-and-priority-tagged" setting;
*) capsman - improved stability when running background scan on CAP;
*) crs3xx - improved maximum allowed ACL rule calculation;
*) crs3xx - improved system stability when creating many ACL rules on CRS317, CRS309, CRS312, CRS326-24S+2Q+ and CRS354 devices;
*) l2tp - improved system stability when processing L2TP control messages;
*) lte - made "no" the default value for "use-network-apn" parameter;
*) lte - made "RG502QEAAAR11A06M4G" the last OTA firmware version update for Chateau 5G in RouterOS 7.1.x release tree;
*) ntp - improved source address usage for reply packets;
*) ospf - fixed default route origination when "default-originate=if-installed" "redistribute" is enabled;
*) ospf - fixed incorrect LSA types when changing area types;
*) ppp - added "comment" option for PPPoE servers;
*) queue - improved system stability when using more than 255 unique packet marks;
*) route - fixed ECMP load balancing in FastPath;
*) route - fixed route addition to VRF from BGP;
*) route - fixed routing configuration export on SMIPS devices;
*) route-filters - renamed "*-set" to "*-list";
*) sfp - improved SFP module detection on CRS106 and CRS112;
*) switch - fixed port-isolation misconfiguration detection when using multiple switches;
*) traffic-flow - do not handle NAT events when "nat-events" is disabled;
*) ups - fixed UPS support;
*) winbox - added "VPN" tab to "Routing/BGP" menu;
*) winbox - added "VRF" parameter for "SSH" and "Telnet" menus;
*) winbox - do not show "Antenna Scan" button on devices that do not support it;
*) wireguard - allow same peer's public key for different interfaces;
*) wireless - added "3gpp-info" parameter to interworking configuration;
*) wireless - added EAP-AKA to interworking's realm configuration;
*) wireless - fixed interface initialization on Metal 2SHPn;

To upgrade, click "Check for updates" at /system package in your RouterOS configuration interface, or head to our download page: http://www.mikrotik.com/download

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 some problem has appeared on device

Please keep this forum topic strictly related to this particular RouterOS release.

msatter · Tue Mar 22, 2022 10:32 am

Thanks, going to try it later today.

I have now 7.1.5 running on my 4011 and the route tables and all the lines that make use of the routes are still there. I am pleased that I can look into using the new features.

It was smooth sailing this updating. First make backup and an export with show-sensitive active, also download the new and the current version of RouterOS.

Thanks for making that possible on request by many, I assume.

:-)

dioeyandika · Tue Mar 22, 2022 10:39 am

RouterBOARD 750G r3, Update from v7.1.3 it break routing table, i dont know if bug or not all my routing table is gone, i restore the routing table with backup it seem okay now, still watching for any bugs

modsx · Tue Mar 22, 2022 10:59 am

In my opinion, the current version of ROS 7 is designed only for arm devices and CPUs with four cores. When placed on an SXT 5HPnD with 400Mhz cpu, its processor often goes 100% down. I don't know what has changed, but old antennas with low-power cpu start to hang even with ROS 6.x.x when exposed to a load (instead of being limited in load, the unit hangs).

zapata · Tue Mar 22, 2022 11:03 am

Why is the below fix still not in stable release?

What's new in 7.2rc2 (2022-Jan-28 11:00):

*) arm - fixed "shutdown" command on hAP ac^2;

haedertowfeq · Tue Mar 22, 2022 11:10 am

break routing table
I am too

blueice · Tue Mar 22, 2022 11:12 am

RouterBOARD 750G r3, Update from v7.1.3 it break routing table, i dont know if bug or not all my routing table is gone, i restore the routing table with backup it seem okay now, still watching for any bugs

i have the same problem to 3 devices so it is a bug. I hope for a bugfix release soon

madejson · Tue Mar 22, 2022 11:17 am

2x hap ac2, 1x hap AC3. No lose/break routing table issue at all.

mseidler · Tue Mar 22, 2022 11:22 am

2x WapAC ok, 1x RB3011 (no firmware update, still 7.1.1) ok

Hominidae · Tue Mar 22, 2022 11:59 am

RouterBOARD 750G r3, Update from v7.1.3 it break routing table

same here for a RB4011, update from 7.1.1

hecatae · Tue Mar 22, 2022 12:19 pm

Why is the below fix still not in stable release?

What's new in 7.2rc2 (2022-Jan-28 11:00):

*) arm - fixed "shutdown" command on hAP ac^2;

There's a couple of fixes that suggest 7.2 is coming to stable soon, the biggest being:

*) lte - made "RG502QEAAAR11A06M4G" the last OTA firmware version update for Chateau 5G in RouterOS 7.1.x release tree;

That's release tree, not stable tree, or stable branch.

fischerdouglas · Tue Mar 22, 2022 12:25 pm

Does anybody is already able to get advertised routes per BGP Peer on v7 stable?

On v7.1.4 release notes, those are the changes that are almost related to that:

*) bgp - fixed VPNv4 route sending to remote peer;
*) route - fixed ECMP load balancing in FastPath;
*) route - fixed route addition to VRF from BGP;
*) route - fixed routing configuration export on SMIPS devices;
*) route-filters - renamed "*-set" to "*-list";

But nothing explicit on advertised routes per peer.

A thread on this forum existis since 2021 November about this:
ROSv7 equivalent of /routing/bgp/advertisements print

@emils Is that on your radar?

irmankurnia · Tue Mar 22, 2022 12:34 pm

This version make my hap ac3 bootloop

Rox169 · Tue Mar 22, 2022 12:36 pm

bad luck, both of my hap ac 3 are working on 7.1.4

mafiosa · Tue Mar 22, 2022 12:40 pm

Currently using v7.99 (internal test image) on RB5009, shall I upgrade? Random crashing issue was not fixed by the test image.

xh116 · Tue Mar 22, 2022 1:02 pm

RB4011 upgraded from 7.13, all routing tables are missing.
Seems the only way is to add manually. (not tested backup/restore

ivicask · Tue Mar 22, 2022 1:08 pm

@strods
viewtopic.php?t=183548#p917086
"pe1chl, msatter, CTassisF, osc86, IntLDaniel, ivicask - Very sorry for missing routing configuration after an upgrade. We are working on a fix for this."
You broke your promise :p

StubArea51 · Tue Mar 22, 2022 1:13 pm

Does anybody is already able to get advertised routes per BGP Peer on v7 stable?

A thread on this forum existis since 2021 November about this:
ROSv7 equivalent of /routing/bgp/advertisements print

@emils Is that on your radar?

This is also a big one for me. It would help get v7 into more widespread use for iBGP and eBGP - especially for peering and transit.

hecatae · Tue Mar 22, 2022 1:26 pm

When will Docker be supported again? T_T T_T T_T T_T

After they fix the bugs for RB5009 and Chateau owners who can only use RouterOS7 on their hardware.

elbob2002 · Tue Mar 22, 2022 1:31 pm

OSPF stopped working - invalid area and invalid instance.

Fix was to change the Router ID in instance from "main" to the IP address of the router.

Jotne · Tue Mar 22, 2022 1:40 pm

When will Docker be supported again? T_T T_T T_T T_T

Wrong thread. This thread is about 7.1.x updates. They will normal only contain bug fixes.
I do also think 7.2Rc is locked for new stuff, so post your request to support@mikrotik.com or here:
viewtopic.php?t=45934

Azma · Tue Mar 22, 2022 2:15 pm

tested on x86, que tree with bridge filter still does not work. SUP-75862.

dave3 · Tue Mar 22, 2022 3:05 pm

Trying to download the update, but it's dialup-modem slow, and keeps disconnecting.

[admin@MikroTik] /system/package/update> download
            channel: stable
  installed-version: 7.1.3
     latest-version: 7.1.4
             status: ERROR: connection disconnected

If I download the npk file from the website and place it in the router's .download directory, will the upgrade work? Or is it a different file it's downloading?

mkx · Tue Mar 22, 2022 3:13 pm

It's the same file ... just download the .npk and put it on root of files folder. Then perform reboot, ROS should pick it up and upgrade. Just make sure architecture is right.

poirus · Tue Mar 22, 2022 3:25 pm

RouterBOARD 750G r3, Update from v7.1.3 it break routing table, i dont know if bug or not all my routing table is gone, i restore the routing table with backup it seem okay now, still watching for any bugs

The same for me on different devices. Roling back to 7.1.3 without restore backup and all routing tables are back

dave3 · Tue Mar 22, 2022 3:29 pm

Thanks, I put the .npk file in the .download directory of the router, and executed the /system/package/update and it disconnected, did it's thing, and rebooted. It's a small file and downloaded instantly over the same connection from the website. Not sure why it won't download properly from the router.

From 7.1.3 to 7.1.4 on an RB750gr3. Seems to be fine. Both ipv4 and ipv6 routing tables look normal and intact. Mine is a pretty simple setup, though.

I did notice that the system/clock timezone settings were missing after the upgrade, though. Comparing an export from before and after the upgrade showed the setting missing. I had to add that back manually.

mafiosa · Tue Mar 22, 2022 3:32 pm

Eagerly waiting for /routing/bgp/advertisements print command equivalent and IPv6 torch functionality.

emils · Tue Mar 22, 2022 4:09 pm

What's new in 7.1.5 (2022-Mar-22 13:03):

*) route - fixed "table" menu emptying after RouterOS upgrade;

anav · Tue Mar 22, 2022 4:13 pm

Wow quick turnaround time............ hard to believe actually, I hope there was some regression testing. ;)

ivicask · Tue Mar 22, 2022 4:14 pm

What's new in 7.1.5 (2022-Mar-22 13:03):

*) route - fixed "table" menu emptying after RouterOS upgrade;

Thank you :)

Amm0 · Tue Mar 22, 2022 4:21 pm

What's new in 7.1.5 (2022-Mar-22 13:03):

*) route - fixed "table" menu emptying after RouterOS upgrade;

I always liked the "testing" -> "stable" -> "long-term" channel scheme in V6. Is there some reason there isn't an intermediate public "testing" channel before stable in V7?

Otherthan I guess "testing" is currently v7.2rc4.... but 7.1.4 does NOT seem to be a promotion of testing as it was in V6.

odge · Tue Mar 22, 2022 4:55 pm

Please see screenshot at

6.47.9 trying to upgrade to 7.1.5

I have 8GB drive with 180Mbytes in use.

Error is

Mikrotik 6.47.9 (long-term)
<identity> Login:
Rebooting....
installed system-7.1.5
ERROR: not enough space for the upgrade

what?

Then i realised... no, this is an OLD x86 Mikrotik that I have had for ages, its the first VM I ever had, its not a CHR, maybe thats why?

Is this upgradable to v7?

Amm0 · Tue Mar 22, 2022 5:03 pm

Please see screenshot at

6.47.9 trying to upgrade to 7.1.5

I have 8GB drive with 180Mbytes in use.

Error is
Mikrotik 6.47.9 (long-term)
<identity> Login:
Rebooting....
installed system-7.1.5
ERROR: not enough space for the upgrade
what?

Then i realised... no, this is an OLD x86 Mikrotik that I have had for ages, its the first VM I ever had, its not a CHR, maybe thats why?

Is this upgradable to v7?

I'd recommend going to the latest stable before moving to V7 anything. 6.47.9 is 2 long-terms (-0.1) ago.

dioeyandika · Tue Mar 22, 2022 5:03 pm

What's new in 7.1.5 (2022-Mar-22 13:03):

*) route - fixed "table" menu emptying after RouterOS upgrade;

wow thats fast :) , i will upgrade in midnight, i will wait other report too, they fix it in half hours that will be a record :)

msatter · Tue Mar 22, 2022 5:35 pm

@strods
viewtopic.php?t=183548#p917086
"pe1chl, msatter, CTassisF, osc86, IntLDaniel, ivicask - Very sorry for missing routing configuration after an upgrade. We are working on a fix for this."
You broke your promise :p

Mikrotik released swiftly a new version that also includes keeping the route table and I have updated my 4011 from 7.1.1 to 7.1.5 now. And it works as before...but with new features.

mkx · Tue Mar 22, 2022 5:42 pm

I always liked the "testing" -> "stable" -> "long-term" channel scheme in V6. Is there some reason there isn't an intermediate public "testing" channel before stable in V7?

Otherthan I guess "testing" is currently v7.2rc4.... but 7.1.4 does NOT seem to be a promotion of testing as it was in V6.

I guess ... MT was somehow forced to push out "stable" v7 so that they (at least on paper) can sell v7-only hardware with some stable version of software on it. BTW, currently tehre isn't a v7 long-term and IMO that's a good thing.
I expect that in a few months, when v7 does stabilise a bit, we will see releases appear in familiar pattern ... e.g. 7.4rc is released into "testing", at the same time 7.3.7 is declared as "stable" and 7.2.12 is made new "long-term" version.

But right now I can imagine developers are in sort of storm trying to get v7 feature set on par with v6, stabilize performance of v7 and add new features every was demanding for the last 10 years. Plus doing some minor maintenance of v6 (in case some nasty bug or vulnerability is discovered). All that at the same time. Not a very pleasant position to be in ...

dave3 · Tue Mar 22, 2022 5:45 pm

I updated from 7.1.4 to 7.1.5. I was just looking at the routing, i.e. /ip/route/print and /ipv6/route/print and it looks the same. I think the "routing table" is something separate, beyond the scope of my knowledge. Hopefully everything is still fine.

Bierkameel · Tue Mar 22, 2022 6:28 pm

RB5009 from 7.1.1 to 7.1.5 No problems so far.

buset1974 · Tue Mar 22, 2022 7:00 pm

Finally i made it to make MPLS L3 works upgrading working MPLS PE with 6.49.5 to 7.1.5
actually it's not perfectly smooth upgrade, i found out that some of MPLS configuration were missing after upgrading to 7.1.5
- MPLS/MPLS Interface were missing
- MPLS/Instance LSR-ID were missing
- MPLS/Instance AFI were uncheck

Just manual fix those configuration and check AFI with "ip" then it's works, i can ping vrf now into other-side PE.

I also find out that OSPF and BGP were perfectly converted/ exported into new v.7.1.5 (no need to make adjustment with OSPF and BGP)

thx

pe1chl · Tue Mar 22, 2022 7:08 pm

I updated from 7.1.4 to 7.1.5. I was just looking at the routing, i.e. /ip/route/print and /ipv6/route/print and it looks the same. I think the "routing table" is something separate, beyond the scope of my knowledge. Hopefully everything is still fine.

The bug in 7.1.4 (which is also present in 7.2rc4...) affects those that have multiple route tables, e.g. in case of loadbalancing/failover with multiple ISPs, or with an overlay network.

armandfumal · Tue Mar 22, 2022 8:57 pm

Does anybody is already able to get advertised routes per BGP Peer on v7 stable?

On v7.1.4 release notes, those are the changes that are almost related to that:
*) bgp - fixed VPNv4 route sending to remote peer;
*) route - fixed ECMP load balancing in FastPath;
*) route - fixed route addition to VRF from BGP;
*) route - fixed routing configuration export on SMIPS devices;
*) route-filters - renamed "*-set" to "*-list";
But nothing explicit on advertised routes per peer.

A thread on this forum existis since 2021 November about this:
ROSv7 equivalent of /routing/bgp/advertisements print

@emils Is that on your radar?

No. It is a big missing today.
check in this page as workaround
https://lostcreek.tech/?p=211

Today in 7.2RC5
*) bgp - added BGP advertisements display (requires output.keep-sent-attributes to be set);
*) routing - added PCAP viewer tool for BGP advertisements debugging purposes;

Tue Mar 22, 2022 9:06 pm

850Gx2 with OSPF+iBGP 7.1.3 -> 7.1.5 ... so far no problems

vmware · Tue Mar 22, 2022 10:04 pm

Please check this video

i have delay action ( delete / disable / Enable ) in firewall CCR2116-12G-4s+ .

https://www.youtube.com/watch?v=ZbkLHiIsys0

But after upgrade to 7.1.5

hecatae · Tue Mar 22, 2022 10:37 pm

Chateau 5G successfully upgraded to 7.1.5,and I'm prevented from upgrading the firmware of the RG502QEA as I'm already at RG502QEAAAR11A06M4G

Guscht · Tue Mar 22, 2022 11:54 pm

CRS326 -> took long -> no problems
CRS309 -> took even longer -> 1 came back online, 1 was dead, after 2 power-off/on it came finally back, but forgot its IP...
hexS -> no problems
mAP lite -> no problems
cAPac -> no problems

in the end, one CRS309 has cost me almost an hour...

SiB · Wed Mar 23, 2022 12:07 am

NetInstall to 7.1.5 the LHGR, done OK.
System Reset configuration to get a Default Configuration, done OK.
Unit was setup by QuickSet - The DHCP Server was ON !!.

Unit disconnect, DHCP not give IP, connection can be done by Mac-Address only.
WinBox Magic of session keep windows open and show uptime status in corner with open log-s too.

Disconnetion of unit is so fast then I not have time to click and correct anything.
I see the DHCP Server is OFF ? Range is missing !. When I try press the DHCP Server ON then I see strange IP RANGE who is show in logs too.

Unit was reset to default by pressing button, and configure manually WITHOUT QUICKSET who probably have got some new BUG as always - Quick Set - this is like always beta feature.

Picture from that story:

VBnQETtBnC.png

ormandj · Wed Mar 23, 2022 12:57 am

I went from 7.1.3 to 7.1.5 and it appears this might have finally fixed all of my 10G SFP+ interfaces flapping on my CRS328-24P-4S+. I haven't seen a port flap in 1+ hours (used to be every 5 minutes). I'll keep monitoring and report back tomorrow, but this is fairly amazing news, if it remains resolved. I've had this issue on every point release since moving to 7.x until now, with 7.1.5 (I did not try 7.1.4).

There was no configuration change to explain this between the upgrades, I merely performed the RouterOS upgrade, then the RouterBOARD firmware update, and that's it. I re-enabled auto-negotiation and all interfaces negotiated 10G, and there hasn't been a port flap since.

@Mikrotik: what changed that would have fixed this? There's nothing in the changelog that looks related. I just want to make sure whatever fixed this also makes it into 7.2.x when it is released!

EDIT: Now all ports are flapping after a few days. This is actually even worse than it was before with just SFP+s flapping.

tamagochi · Wed Mar 23, 2022 1:31 am

RB4011iGS+ 7.1.3->7.1.5 Update seamless.

CCR1009-7G-1C-1S+ 7.1.3->7.1.5 Config crash.
First interface renamed, Wireguard config lost, firewall filter input good but nat table lost. PPP secrets half lost, PPPoe client config lost, IPSEC config lost, Router ID for OSPF lost, dhcp server setup good but fixed leases lost. Certificates lost, users expired.
Routing table ipaddresses good but gateway interface names replaced *number because of lost interfaces, show in red errors (eth1, wireguard, pppoe1 pppoe2)
OSPF interface templates show error and contain bad interfaces.
Config restored in parts from 7.1.3 full export. It looks like it works....

dioeyandika · Wed Mar 23, 2022 1:47 am

okay seem 7.1.5 fix that, but for my problem i added the route table manual before in 7.1.4 so after upgrade to 7.1.5 it have duplicate routing table, then i delete the duplicate one and it mest up again maybe i delete the wrong route table, after fixing it manually it seem fix for now, and seem the invalid mark flag if we add ip route rule is gone now, before after i add some rule in ip route rule it become invalid but after enable and disable the rule it work as it should

brg3466 · Wed Mar 23, 2022 2:04 am

LtAP mini LTE US kit

Still cannot upgrade the modem firmware

[brg3466@LtAP] > /interface/lte/firmware-upgrade lte1 once
installed: R11eL_v05.03.183961
latest: R11eL_v05.04.193841

But once you want to upgrade, it gets stuck because LTE interface disconnected.

I wonder if anyone who successfully upgraded the modem firmware ? I have been trying since 7.1 beta 4, always fail !

gyropilot · Wed Mar 23, 2022 2:11 am

Strange (but thankfully limited) bootlooping happening here on a hAP ac after upgrading to v7.1.5.

I just sent the following bug report through the support website…

——-
I just upgraded a new hAP ac directly from ROS v6.49.5 to v7.1.5. It was running normally on v6.49.5.

Since installing v7.1.5, if I reboot the hAP ac with a Verizon U620L modem inserted in the USB port, it will reboot 3 times in a row (a bootloop x 3) before running normally. The attached autosupout file and log were saved immediately after one of these bootloop events. It is repeatable.

If I reboot the hAP ac WITHOUT the Verizon U620L modem inserted in the USB port, it will reboot only TWO times in a row before running normally. If I then insert the Verizon U620L modem into the USB port (hot insert), the hAP ac reboots itself 3 times in a row before running normally.
——-

Once running, the log shows:

15:51:01 system,error,critical router was rebooted without proper shutdown
15:51:02 system,error,critical kernel failure in previous boot

Amm0 · Wed Mar 23, 2022 2:29 am

I always liked the "testing" -> "stable" -> "long-term" channel scheme in V6. Is there some reason there isn't an intermediate public "testing" channel before stable in V7?

Otherthan I guess "testing" is currently v7.2rc4.... but 7.1.4 does NOT seem to be a promotion of testing as it was in V6.
[...]

But right now I can imagine developers are in sort of storm trying to get v7 feature set on par with v6, stabilize performance of v7 and add new features every was demanding for the last 10 years. Plus doing some minor maintenance of v6 (in case some nasty bug or vulnerability is discovered). All that at the same time. Not a very pleasant position to be in ...

Yeah I wasn't ask for a "long term", more saying it's kinda random right now. Based on a diff of the release notes from 7.1.5 and 7.2rc4, I show:

18 same of the same fixes
10 fixes only that in v7.1.5
20 fixes only in v7.2rc4

So my point is why weren't the 10 fixes only in 7.1.5 not in some rc build first...

Anyway, if anyone is like me deciding if they can move back to the 7.1.x "stable" line from 7.2rc4, here is what I show from diff'ing the two release notes:

        -----------    v7.1.5    -----------                           -----------    v7.2rc4    -----------                   
*) bgp - fixed VPNv4 route sending to remote peer;	      =	*) bgp - fixed VPNv4 route sending to remote peer;
							      >	*) bridge - fixed FastPath when using "frame-types=admit-only
							      >	*) bridge - fixed IP address on untagged bridge interface whe
							      >	*) bridge - fixed PPPoE packet forwarding when using "use-ip-
*) bridge - fixed destination NAT when using "use-ip-firewall =	*) bridge - fixed destination NAT when using "use-ip-firewall
							      >	*) bridge - fixed filter and NAT "set-priority" on ARM64 devi
*) bridge - fixed filter rules when using interface lists;    =	*) bridge - fixed filter rules when using interface lists;
*) bridge - fixed priority tagged frame forwarding when using =	*) bridge - fixed priority tagged frame forwarding when using
*) capsman - improved stability when running background scan  <
							      >	*) console - fixed terminal repainting on F5 and CTRL+L key p
							      >	*) crs3xx - fixed watchdog timer functionality;
*) crs3xx - improved maximum allowed ACL rule calculation;    =	*) crs3xx - improved maximum allowed ACL rule calculation;
*) crs3xx - improved system stability when creating many ACL  =	*) crs3xx - improved system stability when creating many ACL 
*) l2tp - improved system stability when processing L2TP cont =	*) l2tp - improved system stability when processing L2TP cont
							      >	*) lte - added MAC address and IPv6 LL address persistence af
*) lte - made "no" the default value for "use-network-apn" pa =	*) lte - made "no" the default value for "use-network-apn" pa
*) lte - made "RG502QEAAAR11A06M4G" the last OTA firmware ver <
							      >	*) ntp - allow adding duplicate server address if dynamic ent
*) ntp - improved source address usage for reply packets;     <
							      >	*) ntp - fixed multicast mode support;
							      >	*) ntp - improved IPv6 address support;
*) ospf - fixed default route origination when "default-origi =	*) ospf - fixed default route origination when "default-origi
*) ospf - fixed incorrect LSA types when changing area types; <
							      > *) ospf - fixed external LSA not updating after prefix netmas
*) ppp - added "comment" option for PPPoE servers;	      =	*) ppp - added "comment" option for PPPoE servers;
							      >	*) ppp - improved stability when handling large amount of con
*) queue - improved system stability when using more than 255 =	*) queue - improved system stability when using more than 255
*) route - fixed ECMP load balancing in FastPath;	      =	*) route - fixed ECMP load balancing in FastPath;
*) route - fixed route addition to VRF from BGP;	      =	*) route - fixed route addition to VRF from BGP;
*) route - fixed routing configuration export on SMIPS device <
*) route-filters - renamed "*-set" to "*-list";		      =	*) route-filters - renamed "*-set" to "*-list";
*) sfp - improved SFP module detection on CRS106 and CRS112;  <
							      >	*) sms - increased "at-chat" timeout when sending SMS;
*) switch - fixed port-isolation misconfiguration detection w =	*) switch - fixed port-isolation misconfiguration detection w
*) traffic-flow - do not handle NAT events when "nat-events"  <
							      >	*) switch - improved switch chip initialization process on bo
*) ups - fixed UPS support;				      =	*) ups - fixed UPS support;
*) winbox - added "VPN" tab to "Routing/BGP" menu;	      <
							      >	*) vxlan - fixed "group" and "interface" setting reset after 
*) winbox - added "VRF" parameter for "SSH" and "Telnet" menu <
							      >	*) vxlan - fixed running state after reboot when using "inter
							      >	*) wifiwave2 - added "client-isolation" feature;
							      >	*) winbox - added "host-uniq" parameter to PPPoE client inter
*) winbox - do not show "Antenna Scan" button on devices that =	*) winbox - do not show "Antenna Scan" button on devices that
*) wireguard - allow same peer's public key for different int =	*) wireguard - allow same peer's public key for different int
							      >	*) wireguard - fixed IPv6 traffic processing with multiple pe
*) wireless - added "3gpp-info" parameter to interworking con =	*) wireless - added "3gpp-info" parameter to interworking con
*) wireless - added EAP-AKA to interworking's realm configura =	*) wireless - added EAP-AKA to interworking's realm configura
*) wireless - fixed interface initialization on Metal 2SHPn;  <
							      >	*) wireless - correctly preserve WMM priority when receiving 
							      >	*) wireless - improved nv2 link stability;

qatar2022 · Wed Mar 23, 2022 6:58 am

i have the same problem with 4 of RB760iGS,RB960PGS,RB4011iGS+RM,RB5009UG+S+IN,CRS112-8P-4S-IN and CRS328-24P-4S+RM so i think it is a bug
I hope for a bugfix release asap

rushlife · Wed Mar 23, 2022 7:48 am

Hi, installed on ccr1036 and...I have small problem.
High CPU usage. Btw router was doing literally nothing.
Just idling without any usage.

archerious · Wed Mar 23, 2022 7:51 am

Please check this video

i have delay action ( delete / disable / Enable ) in firewall CCR2116-12G-4s+ .

https://www.youtube.com/watch?v=ZbkLHiIsys0

But after upgrade to 7.1.5

I had similar issues on ROS 7.1.3, but usually restarting resolved it.

archerious · Wed Mar 23, 2022 7:56 am

Just updated from ROS 7.1.3 to 7.1.5, and can confirm my routing table is now red, it seems my 2nd ISP isn't working on my CCR2116 post-update.

Hmm.

infabo · Wed Mar 23, 2022 9:27 am

7.1.5 on Chateau and no issues so far.

vmware · Wed Mar 23, 2022 9:36 am

Please check this video

i have delay action ( delete / disable / Enable ) in firewall CCR2116-12G-4s+ .

https://www.youtube.com/watch?v=ZbkLHiIsys0

But after upgrade to 7.1.5
I had similar issues on ROS 7.1.3, but usually restarting resolved it.

I upgrade to 7.1.5 . I reset the factory. But there is still a problem.
😭😭😭😭😭😖😖😖😢

buset1974 · Wed Mar 23, 2022 9:43 am

BGP session for address-family ipv4 vrf xxxxx still not work

SecCon · Wed Mar 23, 2022 11:02 am

Just curious if any of the updates are "high-prio" security related stuff that encourages you to update immediately?

From https://help.mikrotik.com/docs/display/ ... stallation

The package upgrade feature connects to the MikroTik download servers and checks if there is a new RouterOS version for your device.

What are the addresses to those Mikrotik servers?

WebFig > System > Auto Upgrade > Mirrors > Add

pe1chl · Wed Mar 23, 2022 11:15 am

What are the addresses to those Mikrotik servers?

WebFig > System > Auto Upgrade > Mirrors > Add

I think this menu and the functions under it are orphaned. They are not documented and it surprises me that they still are available in v7.
These are probably leftovers from the days there was only one architecture (x86) and a lot more disk space, so you could setup a local mirror of releases.

However, in the normal "check for updates" function the servers are accessed by domain name upgrade.mikrotik.com, which at this moment has addresses 159.148.147.204 and 159.148.172.226.
There is also a name update.mikrotik.com with address 159.148.147.205 but it does not look like it is used much.

SecCon · Wed Mar 23, 2022 11:52 am

Upgrade sources with both those IP addresses for upgrade.mikrotik.com failed. Using my Mikrotik.com creds, not the forum ones. In WebFig 7.1.1 obviously.

Confusing interface

pe1chl · Wed Mar 23, 2022 12:05 pm

As I said, it is an undocumented feature and it probably should not be used anymore.

dg1kwa · Wed Mar 23, 2022 12:38 pm

SFP DOM/DDM still not work on RB760iGS.

junbr0 · Wed Mar 23, 2022 12:46 pm

expecting modularized version release soon. ;)

marlab · Wed Mar 23, 2022 1:03 pm

My RB3011 after factory reset and config import with 7.1.3 had a serious problem with system clock - every appox 3.5 hours the time went backwards by about 30 minutes, which kept triggering watchdog and so system reboot.

With 7.1.5 (both RouterOS and firmware) the problem is still there, just now the clock goes backwards "only" 10 minutes.. :-/

I'm not sure if that can be classified as an improvement...

fragtion · Wed Mar 23, 2022 1:05 pm

I've upgraded several routers (5x chr, 3x RB5009, 1x RB4011, 2x hEX, 2x RB433, 1x LHG-LTE) from v7.1.3 to v7.1.5 without issues
However 1x CRS106-1C-5S lost config on update - Both wireguard interfaces & a wireguard peer were lost. 1 peer remained, but it was corrupted (peer public key AAAAAAAAAAAAAAAz or something)
Fortunately I had a backup, and anydesk to the remote site, otherwise I would have had to physically go to the site to load the backup. So it looks like the issue of config corruption is still ongoing and any device is at risk of losing parts of config on update... Am really hoping this issue is rectified soon so that we can do our upgrades with peace of mind that we're not "rolling the dice" for potential config loss each time, I mean that is just crazy

pe1chl · Wed Mar 23, 2022 1:15 pm

My RB3011 after factory reset and config import with 7.1.3 had a serious problem with system clock - every appox 3.5 hours the time went backwards by about 30 minutes, which kept triggering watchdog and so system reboot.

With 7.1.5 (both RouterOS and firmware) the problem is still there, just now the clock goes backwards "only" 10 minutes.. :-/

What info do you have under System->NTP client and what is shown when you click the Peers button?
It looks like you have a falseticker NTP server in there. When you see a server with a large offset, remove it from the configuration.

pe1chl · Wed Mar 23, 2022 1:18 pm

Fortunately I had a backup,

What does it say in the first posting in this thread and any release thread?

Before an upgrade:
1) Remember to make backup/export files before an upgrade and save them on another storage device;

So it looks like the issue of config corruption is still ongoing and any device is at risk of losing parts of config on update... Am really hoping this issue is rectified soon so that we can do our upgrades with peace of mind that we're not "rolling the dice" for potential config loss each time, I mean that is just crazy

Did you ever start with a fresh netinstall of v7.x and import of a /export (not load of a backup!) on this device?

fragtion · Wed Mar 23, 2022 1:45 pm

Did you ever start with a fresh netinstall of v7.x and import of a /export (not load of a backup!) on this device?

Nope, the device was upgraded from v6.48.x -> v7.1.3 -> v7.1.5 (remotely is the key word here - it's not always easy to do a netinstall on every device)

It shouldn't be necessary to netinstall. There's a clear upgrade path to v7.x from prior versions which is possible & is supposed to work. Upgrading to v7.x should render any previous config (if it must exist) totally inert, and it should have zero possible influence on current resulting operation or stability of the device after the upgrade.

If you say this could have something to do with the way the configuration db migrates, then why must we netinstall to fix that? why can't we just do an /export & /system reset-configuration and restore the config? then the device can rebuild a fresh config db, based on the export file. Surely the same outcome as a netinstall?

Anyway, I'm not convinced that the config loss issue has to do with upgrading from v6, as I had this issue on my RB5009 previously, and that shipped with v7.0.5. So it seems that problem is baked into the current versions of v7 and is still being worked on. Anyway thanks for the suggestions... I'm going to avoid the netinstall because, apart from appearing to be a "shot in the dark" solution that requires a lot of effort to execute, I also actually want to track this issue further and see if/when it occurs again, or gets fixed , so yeah.
Edit: I see you're merely offering that as a suggestion rather than an alleged solution. Great - so let's all keep monitoring and see if we pick up any more patterns here

Jotne · Wed Mar 23, 2022 1:54 pm

Nope, the device was upgraded from v6.48.x -> v7.1.3 -> v7.1.5 remotely is the key word here

This is why you should (if possible) always test on a equal local device before starting on remote device.

pe1chl · Wed Mar 23, 2022 2:00 pm

I'm trying to get clear (from people's findings, including my own) if this is simply a bug in v7 or if it is somehow related to using older versions at first, maybe resulting in a corrupt configuration database.
I persistently lost config on my RB4011 on every reboot and so I netinstalled it to find out. Until now, it has not lost anything but I rebooted it only a couple of times, so will have to see how this develops over longer time.
I agree with you that this is an important problem that really has to be fixed. Hopefully even without netinstall requirement. If necessary, there could be a "repair database" command that effectively exports the config, removes the database and imports it again to do a clean start.

marlab · Wed Mar 23, 2022 2:29 pm

My RB3011 after factory reset and config import with 7.1.3 had a serious problem with system clock - every appox 3.5 hours the time went backwards by about 30 minutes, which kept triggering watchdog and so system reboot.

With 7.1.5 (both RouterOS and firmware) the problem is still there, just now the clock goes backwards "only" 10 minutes.. :-/
What info do you have under System->NTP client and what is shown when you click the Peers button?
It looks like you have a falseticker NTP server in there. When you see a server with a large offset, remove it from the configuration.

Yes, I had the same suspicion and I've checked that. I've tried different NTP public servers, I've also disabled NTP client - it kept doing it regardless of changes I've made.
Under System->NTP client all looks correct, device does the correct sync.

Also yes, the migration to 7.1.3 was as you suggested: factory reset, netinstall, and import previously exported config (not backup/restore).

It looks to me rather like a firmware problem, not RouterOS itself.

loloski · Wed Mar 23, 2022 2:29 pm

Hi MT,

HotSpot functionality with external AP is severely broken on all v7.x MT was able to reproduce the issue but no ETA for the fix. Please give this a priority boost somehow because this is one of the selling features of the device, we can reproduce this on number of device including RB5009 which don't have a workaround because we can't downgrade this to v6

https://help.mikrotik.com/servicedesk/s ... /SUP-74435

Wed Mar 23, 2022 3:38 pm

Hi MT,

HotSpot functionality with external AP is severely broken on all v7.x MT was able to reproduce the issue but no ETA for the fix. Please give this a priority boost somehow because this is one of the selling features of the device, we can reproduce this on number of device including RB5009 which don't have a workaround because we can't downgrade this to v6

https://help.mikrotik.com/servicedesk/s ... /SUP-74435

What exactly is broken ?
Already tried 7.2rc4 ?
There was a hotspot fix in there.
Haven't had the time yet to test myself (but will have to do it shortly for the part I need it for).
But from other users I read their issue was solved using that version.

loloski · Wed Mar 23, 2022 3:45 pm

the verbatim taken directly from my ticket to support

On RB4011 we try to setup a hotspot on wired interface straight from factory reset and put external AP like TP-LINK or any other access point product that we try, the mobile client accessing hotspot after a successful login they are working for a couple of minutes then suddenly will stop to work, we can ping the AP’s IP address but not the RB4011 ip address. we can’t replicate this issue on ROS v6.X only on all v7.x version.

This is consistent and very easy to reproduce just follow the step above

reply from MT

Hello,

Thank you for the report!

We have managed to reproduce the issue locally in our labs and look forward to fixing it on upcoming RouterOS versions, unfortunately, I cannot provide an ETA now.

Wed Mar 23, 2022 3:49 pm

Nope, the device was upgraded from v6.48.x -> v7.1.3 -> v7.1.5 remotely is the key word here
This is why you should (if possible) always test on a equal local device before starting on remote device.

Yes and no.
I also had to start on one occasion really from scratch taking over EXACT config from before the problems (accessing remote LAN using WG was not working). Nothing changed config-wise but starting from scratch it worked.
Some of the brighter minds in this place also had a look at my configs, nobody could indicate a clear error from my side.
Even did a complete diff between both exports before and after clean reset, I couldn't find any apparent reason why it did not work before yet it did after.
So SOMETHING was different in the config starting new which was NOT handled correctly from upgrade to upgrade (and I might be even more worried it is not visible using export, but that's another topic).

And the key part in your response is "if possible".
You can not always do that.
Besides, sooner or later someone has to upgrade remotely for whatever reason and then it is expected those defaults are handled correctly.

loloski · Wed Mar 23, 2022 3:55 pm

Hi MT,

HotSpot functionality with external AP is severely broken on all v7.x MT was able to reproduce the issue but no ETA for the fix. Please give this a priority boost somehow because this is one of the selling features of the device, we can reproduce this on number of device including RB5009 which don't have a workaround because we can't downgrade this to v6

https://help.mikrotik.com/servicedesk/s ... /SUP-74435
What exactly is broken ?
Already tried 7.2rc4 ?
There was a hotspot fix in there.
Haven't had the time yet to test myself (but will have to do it shortly for the part I need it for).
But from other users I read their issue was solved using that version.

yes still broken in 7.2rc4 and if ever it works there we are not going to use this as well because dot1x is also broken with 7.2rcx

BrateloSlava · Wed Mar 23, 2022 5:26 pm

OSPF stopped working after upgrading from 7.1.3 to 7.1.5
By analyzing the text configuration, it was found that one line was "lost" on update. Like this.

/routing id add disabled=no id=172.22.99.254 name=id-slava select-dynamic-id=""

Therefore, the required parameter (router-id) was missing in this configuration. After the fix, everything worked.

/routing ospf instance add name=rt-slava out-filter-chain=ospf-out redistribute=connected router-id=id-slava

TFyre · Wed Mar 23, 2022 5:31 pm

When can you solve BGP+VPLS?

rumahnetmks · Wed Mar 23, 2022 8:02 pm

This version make my hap ac3 bootloop

my hap ac3 update run smooth...

bda · Wed Mar 23, 2022 10:02 pm

Hi, that is true - all router ids lost after update!
Manual recreation needed after update to 7.1.5.
Weird...

OSPF stopped working after upgrading from 7.1.3 to 7.1.5
By analyzing the text configuration, it was found that one line was "lost" on update. Like this.
Code: Select all
/routing id add disabled=no id=172.22.99.254 name=id-slava select-dynamic-id=""
Therefore, the required parameter (router-id) was missing in this configuration. After the fix, everything worked.
Code: Select all
/routing ospf instance add name=rt-slava out-filter-chain=ospf-out redistribute=connected router-id=id-slava

bda · Wed Mar 23, 2022 10:21 pm

Hi, once again about openvpn+udp.
It still does not work on CCR1009 with 7.1.5 Router crashes with any one client session in several seconds after connection.

Dear developers, please fix this issue. it is very important.

ajevremovic · Wed Mar 23, 2022 10:44 pm

LtAP mini LTE US kit

Still cannot upgrade the modem firmware
[brg3466@LtAP] > /interface/lte/firmware-upgrade lte1 once
installed: R11eL_v05.03.183961
latest: R11eL_v05.04.193841

But once you want to upgrade, it gets stuck because LTE interface disconnected.

I wonder if anyone who successfully upgraded the modem firmware ? I have been trying since 7.1 beta 4, always fail !

Try to upgrade device firmware first. My R11e-LTE6 disappeared after upgrading to 7.1.5, and showed up again after upgrading RBM33g's firmware. And I did a remote upgrade from 7.1.3 to 7.1.5... :(

DL7JP · Thu Mar 24, 2022 12:27 am

RB1100AHx4: No problems so far with the update, but also no complex configuration (2 VLANs, WG-Server, and abt 50 FW-Rules).

ormandj · Thu Mar 24, 2022 1:20 am

Update on the link flapping: I still see it every once in a while at 10G with the SFP+s on my CCR2004 and CRS328, but it’s significantly less frequent.

EDIT: Now all ports are flapping after a few days. This is actually even worse than it was before with just SFP+s flapping.

rdlehner · Thu Mar 24, 2022 8:32 am

Updated a RB1100AHx4 yesterday evening from 7.1.3 to 7.1.5
Configuration messed up, lost network connectivity. Bridges were gone, most of the IP addresses (except wireguard IPs), and routing table all unassigend/red.
Had to restore previously backed up configuration.

infabo · Thu Mar 24, 2022 2:28 pm

hAP lite upgraded fine. 7.1.3 -> 7.1.5. The upgrade took like > 10min and the device be up again. But it finished fine.

sirbryan · Thu Mar 24, 2022 11:28 pm

Something that has changed about OSPF in 7.2RC4/7.1.4/7.1.5, when loaded onto one of my CCR2004's (lab connection to production network), sends an announcement via OSPF to all the other routers, causing one of the cores to redirect traffic to itself. I can't keep it up long enough to determine what it is because it takes everybody down.

The above release notes talk about "default route" "if-installed", but this router doesn't announce the default route at all, and all other routers it peers with are on 7.1.3 (5009's) or 6.47.10.

This router does have a default route on different VRFs for some hotspot VLANs to other providers, but it's not running OSPF on those VRFs at all.

vacari · Fri Mar 25, 2022 2:05 am

RB3011UiAS-RM
Upgrade 7.1.3 -> 7.1.5
BootLoop
No Netinstall access via ETH1 port (+-20sec pressing reset)
Any suggestion? Console cable indication? Which one?
I'm kind of desperate. Thank you!

vacari · Fri Mar 25, 2022 2:44 am

BootLoop
No Netinstall access via ETH1 port (+-20sec pressing reset)

Hey, guys,
I was able to access through Netinstall netinstall-6.49.5 and gained access again.
I'm going all environment to version 6 for a few years
Thank you anyway.

Jotne · Fri Mar 25, 2022 7:18 am

Upgraded a hEX RB750gv3 to 7.1.5 everything seems to went fine except when I look at System->Package
In check for Updates it only show ERROR: connection timed out when I look for new version.
Same for long term, stable, testing and development.
I can ping mikrotik.com from terminal, so internet connection is ok.

dave3 · Fri Mar 25, 2022 7:34 am

I had a similar problem on my RB750gr3 with connection timeouts from system->package. I don't know why it's like that. It would try to download (at dial-up modem speed or worse) and constantly timeout and restart. Eventually, it did download, very slowly. But also, I just downloaded the npk file from the website and uploaded it to the router to do the update. That worked, and the download of the small file from the website was instant.

thuety · Fri Mar 25, 2022 8:44 am

Updated from 7.1.3 to 7.1.5 without any issues:
- RB2011UiAS-2HnD
- RB941-2nD-TC

nostromog · Fri Mar 25, 2022 9:35 am

Upgraded a hEX RB750gv3 to 7.1.5 everything seems to went fine except when I look at System->Package
In check for Updates it only show ERROR: connection timed out when I look for new version.
Same for long term, stable, testing and development.
I can ping mikrotik.com from terminal, so internet connection is ok.

I have been seeing this for a long time: I can download from a machine natted to the router very fast. I think it is due to this:

[boss@MT] > :ping download.mikrotik.com
  SEQ HOST                                     SIZE TTL TIME       STATUS                                                          
    0 159.148.172.226                                              timeout                                                         
    1 159.148.172.226                                              timeout                                                         
    2 159.148.172.226                                              timeout                                                         
    3 159.148.172.226                                              timeout                                                         
    sent=4 received=0 packet-loss=100% 
[boss@MT] > :ping download.mikrotik.com
  SEQ HOST                                     SIZE TTL TIME       STATUS                                                          
    0 159.148.147.204                            56  47 75ms458us 
    1 159.148.147.204                            56  47 75ms275us 
    sent=2 received=2 packet-loss=0% min-rtt=75ms275us avg-rtt=75ms366us max-rtt=75ms458us

Apparently the machines are trying to use download.mikrotik.com and it resolves to 2 different IPs, one of them is offline. My laptop is clever enough to "blacklist" the wrong one, but mikrotik doesn't.

Jotne · Fri Mar 25, 2022 10:01 am

Strange, it works now after some time.

But this is a BUG for MT to fix.

In stable channel, it shows 7.1 even if I have 7.1.5. So according to MT, I should downgrade???

pe1chl · Fri Mar 25, 2022 10:37 am

It can point to some MTU problem further upstream. E.g. you have the MikroTik behind an ISP router that has a PPPoE link with MTU 1492 or so, but on the LAN side the MTU is 1500 and that is also what the MikroTik sees.
The whole thing is relying on PMTUD then, which is broken for the MikroTik download servers.
But in this case there often are other random sites that are slow or don't work at all.

sirbryan · Fri Mar 25, 2022 12:09 pm

Something that has changed about OSPF in 7.2RC4/7.2RC5/7.1.5, when loaded onto one of my CCR2004's (lab connection to production network), sends an announcement via OSPF to all the other routers, causing one of the cores to redirect traffic to itself.

I found the problem.

If I load anything above 7.1.3 (7.2rc4/5, 7.1.5) on my CCR2004-1G-12S+2XS, my OSPF out filters are ignored, specifically the reject statements meant to keep all but a couple of subnets from leaving the router. Routes learned from other routers were further propagated, causing some of the core routers to circularly reference each other.

Turns out that router and another CCR2004 running 6.47.10 each have a second OSPF instance sharing only each other's routes. I changed that instance's area to a different area ID, then changed it to a stub. It stopped the circular references but it didn't stop the private routes from leaking. That's how I noticed the reject rules weren't working.

add chain=ospf-out disabled=no rule="if (dst in 100.64.0.0/10 && dst-len in 24-32) { accept; }"
add chain=ospf-out disabled=no rule="if (dst in 0.0.0.0/0 && dst-len in 0-32) { reject; }"

This has worked all the way up to 7.1.3. It doesn't work on 7.2rc4+ or 7.1.5.

On another note, an upgrade from 7.1.3 to 7.1.5 on an RB5009 failed. It shows up in its neighbors' entries as having 7.1.5, but it's not responding on its IP's and I can't mac-telnet to it from any neighboring device. (It's out in the shop and it's 4:00 AM. I'll go out and look at it later...)

jult · Fri Mar 25, 2022 2:23 pm

Our Metal 52c access point kicks me out of Winbox after a 3 seconds, winbox keeps reconnecting and being kicked out. This is ever since 7.1.4 was put on it. I can no longer access the device, because I switched off its web-UI server. I used to be able to connect after having power-cycled, but now this workaround no longer works.
I would have to physically reset the device manually, or is there another way?

Fri Mar 25, 2022 2:42 pm

Our Metal 52c access point kicks me out of Winbox after a 3 seconds, winbox keeps reconnecting and being kicked out. This is ever since 7.1.4 was put on it. I can no longer access the device, because I switched off its web-UI server. I used to be able to connect after having power-cycled, but now this workaround no longer works.
I would have to physically reset the device manually, or is there another way?

Which version of Winbox are you using ?
Try the latest.

irmankurnia · Fri Mar 25, 2022 2:52 pm

still memory leak on my hap ac3 after few days and rebooting itself

Fri Mar 25, 2022 3:22 pm

still memory leak on my hap ac3 after few days and rebooting itself

I'm not seeing that ?

dave3 · Fri Mar 25, 2022 4:39 pm

It can point to some MTU problem further upstream. E.g. you have the MikroTik behind an ISP router that has a PPPoE link with MTU 1492 or so, but on the LAN side the MTU is 1500 and that is also what the MikroTik sees.
The whole thing is relying on PMTUD then, which is broken for the MikroTik download servers.
But in this case there often are other random sites that are slow or don't work at all.

For what it's worth, I'm connected to the ISP ONU with DHCP, MTU=1500, and I see the same timeouts and slow download.

[admin@MikroTik] > ping download.mikrotik.com
  SEQ HOST                                     SIZE TTL TIME       STATUS
    0 159.148.172.226                                              timeout
    1 159.148.172.226                                              timeout
    2 159.148.172.226                                              timeout
    3 159.148.172.226                                              timeout
    sent=4 received=0 packet-loss=100%

[admin@MikroTik] > ping download.mikrotik.com
  SEQ HOST                                     SIZE TTL TIME       STATUS
    0 159.148.147.204                            56  51 264ms937us
    1 159.148.147.204                            56  51 266ms147us
    2 159.148.147.204                            56  51 269ms439us
    3 159.148.147.204                            56  51 264ms230us
    sent=4 received=4 packet-loss=0% min-rtt=264ms230us avg-rtt=266ms188us max-rtt=269ms439us

pe1chl · Fri Mar 25, 2022 4:55 pm

ping is not a useful diagnostic for such things, it may well be that the server does not allow ping but still works for download.
it may be there is a network issue at MikroTik at the moment though, I noticed that help.mikrotik.com is down on IPv6 but working on IPv4.

herger · Fri Mar 25, 2022 5:35 pm

Updated a CRS328 from 7.1.2 to 7.1.5 and i don't see the Mikrotik SNMP MIB answered anymore ): Means i have a big red exclamation mark in my monitoring. Is this the same for all devices running 7.1.5? Asking for a friend, as it might hinder all his update plans (;

jult · Fri Mar 25, 2022 5:39 pm

Our Metal 52c access point kicks me out of Winbox after a 3 seconds, winbox keeps reconnecting and being kicked out. This is ever since 7.1.4 was put on it. I can no longer access the device, because I switched off its web-UI server. I used to be able to connect after having power-cycled, but now this workaround no longer works.
I would have to physically reset the device manually, or is there another way?
Which version of Winbox are you using ?
Try the latest.

It is the latest (.35). I got in using ssh, luckily I left that one open/running. That way I did a factory reset without having to climb the tower here.. Still, my config was nothing out of the ordinary, and it's a little worrying that even "long-term" versions aren't truly stable. I'm wondering to what RoS and bootfw version I should downgrade this device.

pe1chl · Fri Mar 25, 2022 6:05 pm

Well, your doubt has been resolved by MikroTik, they have declared 7.1.5 the "long term" version, so no need to downgrade anymore.
(of course this will not solve any problems you may have with 7.1.5)

Remember the release channel (long-term, stable, development) only defines the release schedule of the software, not the reliability. "stable" versions often are released without any real testing, and fixed versions quickly follow eachother. It would be different when some development version after testing time was promoted to "stable", and some stable version that turns out to be of good quality is then later promoted to long-term, but that is not how it works at MikroTik.

garlicbulb · Fri Mar 25, 2022 9:31 pm

Upgraded 7.1.3 to 7.1.5 on my RB751U-2Hnd

System is unstable for first 10 minutes or so. CPU is 100% and attempts to logon result in crash. After the time, CPU goes back up normal. This was also true on 7.1.3.

My RB952Ui-5ac2nD seems to be stable on 7.1.3

cjmaxik · Fri Mar 25, 2022 10:41 pm

Updating firmware on my hAP ac from 7.1.3 to 7.1.5 wiped the config entirely... Again, each update gives more headaches than solves. This is it. Thank you, Mikrotik, this was my last piece of gear from you.

infabo · Sat Mar 26, 2022 9:37 am

Easy fix i would day?

keraia · Sat Mar 26, 2022 3:17 pm

Did anyone try version 7.1.x stable on RB922UAGS ? How did the upgrade from 6.49 go?

pe1chl · Sat Mar 26, 2022 4:12 pm

RB922UAGS has 128M of Flash so you can partition it (2 partitions), copy your existing install, and have a go. When it fails, just switch back to the old partition.

kehrlein · Sat Mar 26, 2022 6:44 pm

Update of RB760iGS (hEX S) led to inaccessibility of the device; netinstall solved the issue.

Updated soft- and firmware on these models without any issues (besides RoMON still not working since 7.1.x):
RBcAPGi-5acD2nD (cAP ac)
RB750GL
CRS309-1G-8S+

Amm0 · Sat Mar 26, 2022 8:58 pm

NetInstall to 7.1.5 the LHGR, done OK.
System Reset configuration to get a Default Configuration, done OK.
Unit was setup by QuickSet - The DHCP Server was ON !!.

Unit disconnect, DHCP not give IP, connection can be done by Mac-Address only.
WinBox Magic of session keep windows open and show uptime status in corner with open log-s too.

Disconnetion of unit is so fast then I not have time to click and correct anything.
I see the DHCP Server is OFF ? Range is missing !. When I try press the DHCP Server ON then I see strange IP RANGE who is show in logs too.

Unit was reset to default by pressing button, and configure manually WITHOUT QUICKSET who probably have got some new BUG as always - Quick Set - this is like always beta feature.

It's done this in V7 a lot, it creates DHCP network 0.0.0.0/0. It will then create the right one for the subnet if you hit the apply again, e.g. "twice". BUT still want to make sure DHCP Server is still check, since sometimes the "first apply" seemingly doesn't apply the DHCP Server setting. The 0.0.0.0 is never cleaned up, even if you re-apply.

I only use it to get a default firewall and some base config that my scripts tweak further (e.g. easier to know it has a firewall, internet, etc. before doing VLANs, multiple WANs, etc...). But boy there is always something it doesn't quite set right. And what that is does vary by version. The DHCP Server 0.0.0.0 thing with QuickSet has been around in a few builds now.

NOW I think this only happens if you change the IP address it's using, especially if it showed 0.0.0.0 as IP at anypoint. Basically I think QuickSet creates the DHCP Server BEFORE applying the IP address you set in QuickSet. Using the value that was displayed BEFORE you changed it.

This is what one run of QuickSet leaves in v7.1.5, at least on an RB953 I just tested:

/ip dhcp-server
add address-pool=dhcp interface=bridge name=dhcp1
/ip dhcp-server network
add address=0.0.0.0/24 gateway=0.0.0.0 netmask=24

A feature request here be some thing to VIEW the quickset script, since there are multiple "Profiles", some thing like:

 /system/default-configuration/quick-set/print

Jotne · Sat Mar 26, 2022 11:56 pm

I do have a memory leak after upgrading a 750Fr3 (that have a large config) from 6.49.2->6.49.5->7.1.5 friday morning.
After the upgrade memory usage only goes up. I will lett it go some week and follow it closely. If it does not stop, I will try to stop DoH that I have had problem with before.
.

Memory 7.1.5.png

YO3IPT · Sun Mar 27, 2022 1:41 am

Something that has changed about OSPF in 7.2RC4/7.2RC5/7.1.5, when loaded onto one of my CCR2004's (lab connection to production network), sends an announcement via OSPF to all the other routers, causing one of the cores to redirect traffic to itself.
I found the problem.

If I load anything above 7.1.3 (7.2rc4/5, 7.1.5) on my CCR2004-1G-12S+2XS, my OSPF out filters are ignored, specifically the reject statements meant to keep all but a couple of subnets from leaving the router. Routes learned from other routers were further propagated, causing some of the core routers to circularly reference each other.

Turns out that router and another CCR2004 running 6.47.10 each have a second OSPF instance sharing only each other's routes. I changed that instance's area to a different area ID, then changed it to a stub. It stopped the circular references but it didn't stop the private routes from leaking. That's how I noticed the reject rules weren't working.
Code: Select all
add chain=ospf-out disabled=no rule="if (dst in 100.64.0.0/10 && dst-len in 24-32) { accept; }"
add chain=ospf-out disabled=no rule="if (dst in 0.0.0.0/0 && dst-len in 0-32) { reject; }"
This has worked all the way up to 7.1.3. It doesn't work on 7.2rc4+ or 7.1.5.

On another note, an upgrade from 7.1.3 to 7.1.5 on an RB5009 failed. It shows up in its neighbors' entries as having 7.1.5, but it's not responding on its IP's and I can't mac-telnet to it from any neighboring device. (It's out in the shop and it's 4:00 AM. I'll go out and look at it later...)

I have noticed this as well, OSPF fails, cycling from Full state down to Init - caused by redistributed routes that should have been rejected by the output filter. Another thing I have noticed is that although output filters are ignored, applying the same rule to the input filter (on the opposite end) seems to work ok.
My setup is a basic one, with a RB4011 and a couple of hAP ac2's in a broadcast network with a single backbone area. I imagine for more complex setups there might be more problems...

accarda · Sun Mar 27, 2022 10:41 am

Since Ros 7.x.x I see this bug with file deletion and what is displayed in Winbox.
I have a backup script that creates files "some-name_date" and saves to disk.
Then each week the previous file is deleted before creating the new set for upload to FTP server.
If I'm already logged with Winbox, when I open Files I will see both set of files: the old and new.
But in reality the old set was properly removed, but it's still visible like if it was not removed.
As soon as I logout and login with Winbox the old files are no longer visible.
I can see this on all my devices with different disk types, so it's a general bug.

indy · Sun Mar 27, 2022 11:38 am

Excessive CPU usage on the RB3011 is still not fixed.
Does anyone know if Mikrotik has at least recognized the problem at some point?

jult · Sun Mar 27, 2022 12:59 pm

I would like to stress that upgrading to 7.* should not be done for everyone. Thus far only the versions the devices came with seem to be stable for me.

pe1chl · Sun Mar 27, 2022 2:05 pm

Since Ros 7.x.x I see this bug with file deletion and what is displayed in Winbox.

I think in general v7 is less agressive at refreshing window contents in winbox. You can press F5 to refresh it at any time. In v6 when something changes in the router it would also change in the winbox window, but in v7 that generally does not happen. This may mean there is less network traffic for an open winbox and also probably less load on the router.
I would prefer if it would be an option to get everything refreshed. I keep winbox open on some important routers with a couple of windows with status information, and they no longer auto-update on v7, where that works OK on v6.

Amm0 · Sun Mar 27, 2022 4:23 pm

Since Ros 7.x.x I see this bug with file deletion and what is displayed in Winbox.
I think in general v7 is less agressive at refreshing window contents in winbox. You can press F5 to refresh it at any time. In v6 when something changes in the router it would also change in the winbox window, but in v7 that generally does not happen.

This is totally true. I never even knew when winbox had a refresh before V7, it just updated. This is especially in IP>Routes with stuff like recursive routes, where you have both the actual delay in recalculating the route table AND winbox getting that calculation. If you don't hit F5, you might think some recent route configuration was misconfigured – instead it may be just winbox has updated IP>Routes with the new state.

But "refresh feature" wasn't really needed before. I'm hopeful as V7 bugs are fixed, and eventually optimized more, F5 and refresh won't be a long term need.

Right now, hit F5 if ever in doubt if something happen. It likely did and winbox just ain't showing it.

moiskey · Sun Mar 27, 2022 10:16 pm

Greetings. i lost compatibility with my old network card in mikroitk x86 (qlogic bmc5709c).
When configuring, it is simply with the status of the network cable disconnected (it does not change the status of running).
I hope it can be solved

oe5nip · Mon Mar 28, 2022 7:38 am

I have installed v7.1.5 and also 7.1.3

There is a bug I think.

I have a Connection to a router via AP.

This stopped working after update.

After rebounding all is ok.

pe1chl · Mon Mar 28, 2022 11:42 am

Impossible to work with such reports when the previous version is not mentioned and the config export is not included.

BillyVan · Mon Mar 28, 2022 11:45 am

On Chateau 5G after update 7.1.5

infabo · Mon Mar 28, 2022 12:21 pm

Please read the changelog:

*) lte - made "RG502QEAAAR11A06M4G" the last OTA firmware version update for Chateau 5G in RouterOS 7.1.x release tree;

mthqwork · Mon Mar 28, 2022 12:50 pm

I have upgraded my 7.1 CHR instance to 7.1.5. Regret it.
After that, device did not came back online anymore, console shows:

Load system

ERROR: could not mount disk!
Please attach it somewhere else.

hecatae · Mon Mar 28, 2022 1:42 pm

On Chateau 5G after update 7.1.5

working as intended.

BillyVan · Mon Mar 28, 2022 2:01 pm

On Chateau 5G after update 7.1.5
working as intended.

For checking only the message could be just you have the latest version

ormandj · Mon Mar 28, 2022 3:09 pm

Is anybody else now seeing port flapping on regular ethernet ports on CRS328-24P-4S+? Before I was seeing every 5-minute flaps on SFP+s at 10G, now I am seeing flaps on all ports (moving from 7.1.3 to 7.1.5). The frequency is less, but the impact is wider.

Screen Shot 2022-03-28 at 07.08.03.png

With more research, it seems to be associated with disk activity (graphing writes, etc.) I've disabled as much as I can, and it seems to have helped, but this is clearly still a major problem.

Felixhuynh271 · Mon Mar 28, 2022 5:52 pm

hi I'm new to Ros. I own a rb4001 and heard that bridge hw offloading is supported on rb4011 since OS 7.x, but when I enabled hw offloading in bridge (while using vlan filtering) the hw offload status still greyed out in each port. Is the functions integrated yet or am I missing something? Thanks in advance

pe1chl · Mon Mar 28, 2022 6:10 pm

I think you need to turn off STP on the bridge to enable hw offloading in the 4011.

benlg · Mon Mar 28, 2022 6:32 pm

At least it works here with RSTP, and just tested with STP, it also works.
Be sure you only have one bridge, as only the first bridge will be hardware-offloaded.

Felixhuynh271 · Mon Mar 28, 2022 6:59 pm

for protocol I tested with: none, stp and rstp the hw offload still greyed out. I did try to enable it by CLI this is what it shown:
[admin@Mikrotik RouterOS] > /interface/ethernet/switch set 0 l3-hw-offloading=yes
failure: L3 HW Offload not supported

mkx · Mon Mar 28, 2022 7:06 pm

failure: L3 HW Offload not supported

L3 offloading is not supported on RB4011 ... L2 however is. That means that bridging VLAN infested traffic within same VLAN can be entirely handled by switch chips (if ports are connected to same switch chip). This wasn't the case in v6.

mkx · Mon Mar 28, 2022 7:10 pm

... when I enabled hw offloading in bridge (while using vlan filtering) the hw offload status still greyed out in each port.

Beware that only single bridge can offload to switch chip. If you have more than one bridge (some people like it that way), then you can force ROS to offload a particular bridge ... with some trickery.

Felixhuynh271 · Mon Mar 28, 2022 7:11 pm

oh I might have misread the changelog then. Does the greyed out hw offload in bridge>ports apply to l3 or l2 hw-offloading? And does l2 hw-offloading need to be enable or is it enabled by default? Thanks

mkx · Mon Mar 28, 2022 7:15 pm

That status is about L2. It is enabled by default for bridge, can be toggled per port. If all bridge ports have hw unset, bridge will not be HW offloaded (obviously).

And don't forget: only ports, run by switch chip, can be HW offloaded. That excludes, for example, wireless interfaces (configured as bridge ports).

Felixhuynh271 · Mon Mar 28, 2022 7:27 pm

Hi, thank you for your quick replies. Apparently, hw offloading will not be enable with either IGMP snooping or DHCP snooping enable which is quite weird

mkx · Mon Mar 28, 2022 7:47 pm

It's not that weird. In order for snooping to work, switch (device is being used like one) needs to process certain types of packets. Switch chips with longer list of features can either process those packets autonomously ar at least pick packets and only forward to CPU those that need processing. Low-end switch chips don't have such capabilities which means all packets have to pass CPU if something "complex" needs to be done.

Felixhuynh271 · Mon Mar 28, 2022 7:57 pm

got it thank you all

hecatae · Mon Mar 28, 2022 9:56 pm

working as intended.
For checking only the message could be just you have the latest version

I don't disagree, I'm using 7.2rc5 on my chateau.

sirbryan · Mon Mar 28, 2022 10:34 pm

I found the problem.

If I load anything above 7.1.3 (7.2rc4/5, 7.1.5) on my CCR2004-1G-12S+2XS, my OSPF out filters are ignored, specifically the reject statements meant to keep all but a couple of subnets from leaving the router. Routes learned from other routers were further propagated, causing some of the core routers to circularly reference each other.
Code: Select all
add chain=ospf-out disabled=no rule="if (dst in 100.64.0.0/10 && dst-len in 24-32) { accept; }"
add chain=ospf-out disabled=no rule="if (dst in 0.0.0.0/0 && dst-len in 0-32) { reject; }"
This has worked all the way up to 7.1.3. It doesn't work on 7.2rc4+ or 7.1.5.
I have noticed this as well, OSPF fails, cycling from Full state down to Init - caused by redistributed routes that should have been rejected by the output filter. Another thing I have noticed is that although output filters are ignored, applying the same rule to the input filter (on the opposite end) seems to work ok.
My setup is a basic one, with a RB4011 and a couple of hAP ac2's in a broadcast network with a single backbone area. I imagine for more complex setups there might be more problems...

Thanks for confirming this. Support seemed to think I have it set up wrong. I quoted the current documentation for RouterOS 7 regarding output filters, and I got this back:

"Behaviour changed due to attempts to fix problems with default route distribution. We will try to change the behaviour for upcoming versions similar as it was in ROS v6."

I've got a dozen routers with four in the core (CCR1036, RB5009, CCR2004), two facing my upstream providers (CCR2004 & RB4011), and five customer-facing edge routers (RB4011) in a ring. All but the core ARM64 routers run RouterOS 6.

I upgraded from 7.1.3 to 7.1.5 on the core CCR2004 and the 1036 and 5009's kept throwing the traffic around in circles (based on traceroutes from behind the 2004). I backed the 2004 out to 7.1.3, then tried it on one of the 5009's and the exact same thing happened.

I've been trying to point out to support that a working config on three routers on 7.1.3 does not work on 7.1.5 nor 7.2rc4/5, and that in fact it breaks things horribly, but I keep getting told I'm doing it wrong...

ksteink · Tue Mar 29, 2022 1:19 am

I did upgrade a hEX S and no problems. I tried a CRS326 and no matter what I did I got not enough space for upgrade.

I tried this:
- Update via /System/Package/update menu
- Manual update by uploading the ROS main package for ARM.
- I tried an older version of ROS (v.7.1.2) that is a bit smaller and nothing
- I removed the ppp, wireless, mpls and hotspot packages and same error.

I see 27% free space (But even I upload any file manually) the 27% never changes so there is something wrong showing the available space.

Both devices were in 6.49.5 version and they are remote (I cannot do a Netinstall and I don't want to go that route either).

Any other suggestions?

YO3IPT · Tue Mar 29, 2022 1:51 am

Regarding the OSPF Output filter problem for redistributed connected routes, I discovered a fix / workaround.
If you uncheck the redistributed connected routes checkbox, but leave the OSPF output filter selected - it will work correctly. I think the checkbox overrides the filter and redistributes all connected routes no matter what.
Before 7.1.5 it worked correctly, I had both the checkbox selected and the proper output filter rules.
Maybe this will help fix some other issues with OSPF.

syahjo · Tue Mar 29, 2022 4:52 am

Hi all, perhaps any one can share some experiences or maybe admin/vendor or else can help.
Will be appreciate gracefully.

Didn't know it's a bug or just new scheme feature, Get different ip address
the setup was simple & default, only standard hotspot captive portal
RB3011

some new user can't access internet after successfully logon :((

example : user got ip address from /ip/dhcp server/ leases 192.168.88.221 and have ip address from /ip/hotspot/active 192.168.88.218
with the same mac addresses

double ip.jpg

here the set

# mar/23/2022 10:52:33 by RouterOS 7.1.4
# software id = ****
#
# model = RouterBOARD 3011UiAS
# serial number = 7****
/interface bridge
add admin-mac=6*********E auto-mac=no comment=defconf name=bridge
/interface ethernet
set [ find default-name=ether1 ] speed=100Mbps
set [ find default-name=ether2 ] name=ether2-master speed=100Mbps
set [ find default-name=ether3 ] speed=100Mbps
set [ find default-name=ether4 ] speed=100Mbps
set [ find default-name=ether5 ] speed=100Mbps
set [ find default-name=ether6 ] name=ether6-master speed=100Mbps
set [ find default-name=ether7 ] speed=100Mbps
set [ find default-name=ether8 ] speed=100Mbps
set [ find default-name=ether9 ] speed=100Mbps
set [ find default-name=ether10 ] speed=100Mbps
set [ find default-name=sfp1 ] advertise=\
    10M-half,10M-full,100M-half,100M-full,1000M-half,1000M-full

/interface list
add exclude=dynamic name=discover
/interface wireless security-profiles
set [ find default=yes ] supplicant-identity=MikroTik
/ip hotspot profile
add dns-name=hot.spot hotspot-address=192.168.88.1 name=hsprof2
/ip hotspot user profile
set [ find default=yes ] shared-users=10
/ip pool
add name=default-dhcp ranges=192.168.88.10-192.168.88.254
/ip dhcp-server
add add-arp=yes address-pool=default-dhcp interface=bridge name=defconf
/ip hotspot
add address-pool=default-dhcp addresses-per-mac=unlimited disabled=no \
    interface=bridge name=server1 profile=hsprof2
/ip hotspot user profile
add address-pool=default-dhcp name=4m rate-limit=4m/4m shared-users=100
add address-pool=default-dhcp name=10m rate-limit=10m/10m shared-users=100
add address-pool=default-dhcp name=3m rate-limit=3m/3m shared-users=100
/port
set 0 name=serial0
/snmp community
set [ find default=yes ] addresses=0.0.0.0/0
/user group
set full policy="local,telnet,ssh,ftp,reboot,read,write,policy,test,winbox,pas\
    sword,web,sniff,sensitive,api,romon,dude,tikapp,rest-api"
/zerotier
set zt1 comment="ZeroTier Central controller - https://my.zerotier.com/" \
    disabled=yes disabled=yes name=zt1 port=9993
/interface bridge port
add bridge=bridge comment=defconf interface=ether2-master
add bridge=bridge comment=defconf interface=ether6-master
add bridge=bridge comment=defconf hw=no interface=sfp1
add bridge=bridge interface=ether3
add bridge=bridge interface=ether4
add bridge=bridge interface=ether5
add bridge=bridge interface=ether7
add bridge=bridge interface=ether8
add bridge=bridge interface=ether9
add bridge=bridge interface=ether10
/ip neighbor discovery-settings
set discover-interface-list=discover
/ipv6 settings
set disable-ipv6=yes
/interface detect-internet
set detect-interface-list=all
/interface list member
add interface=ether2-master list=discover
add interface=ether3 list=discover
add interface=ether4 list=discover
add interface=ether5 list=discover
add interface=sfp1 list=discover
add interface=ether6-master list=discover
add interface=ether7 list=discover
add interface=ether8 list=discover
add interface=ether9 list=discover
add interface=ether10 list=discover
add interface=bridge list=discover
/ip address
add address=192.168.88.1/24 comment=defconf interface=bridge network=\
    192.168.88.0
/ip cloud
set ddns-enabled=yes
/ip dhcp-client interface=ether1

/ip dhcp-server lease
add address=192.168.88.254 client-id=1:14:***************  mac-address=14:************ server=\
    defconf
add address=192.168.88.244 client-id=1:14:***************  mac-address=14:***************** server=\
    defconf
add address=192.168.88.248 client-id=1:**************78 mac-address=\
    B4:**************8 server=defconf
add address=192.168.88.237 client-id=1:****************d comment=cctv \
    mac-address=***************:8D server=defconf
/ip dhcp-server network
add address=192.168.88.0/24 comment=defconf gateway=192.168.88.1
/ip dns
set allow-remote-requests=yes servers=192.168.88.1
/ip firewall nat
add action=passthrough chain=unused-hs-chain comment=\
    "place hotspot rules here" disabled=yes
add action=masquerade chain=srcnat comment="defconf: masquerade" \
    out-interface=ether1
add action=masquerade chain=srcnat comment="masquerade hotspot network" \
    src-address=192.168.88.0/24

/ip hotspot user
add name=s6421 server=server1

Jotne · Tue Mar 29, 2022 8:27 am

As seen form this post viewtopic.php?p=921640#p921640
hEX RB750Gr3 continues to leak memory since upgrade to 7.1.5
.

Memory 7.1.5.png

Tue Mar 29, 2022 5:37 pm

You're not going to like it ...
Since that jump is rather high (even with the intermediate stop), upgrade to 7.1.5 with CLEAN config.
Then take over config from previous export line by line, carefully paying attention to what's already present in default settings.

sirbryan · Tue Mar 29, 2022 6:05 pm

Regarding the OSPF Output filter problem for redistributed connected routes, I discovered a fix / workaround.
If you uncheck the redistributed connected routes checkbox, but leave the OSPF output filter selected - it will work correctly. I think the checkbox overrides the filter and redistributes all connected routes no matter what.
Before 7.1.5 it worked correctly, I had both the checkbox selected and the proper output filter rules.
Maybe this will help fix some other issues with OSPF.

That's what support told me was going on and to disable all redistribution checkboxes. But even after doing that I still ran into problems. I probably should test it again at 2:00 a.m. one of these days.

npeca75 · Tue Mar 29, 2022 9:18 pm

I did upgrade a hEX S and no problems. I tried a CRS326 and no matter what I did I got not enough space for upgrade.

If i remember correctly, on ARM devices only way to upgrade is from bundled system
other platforms work well, v6 bundled package or v6 individual packages upgrading to v7
but ARM , nope ...

so i am stuck as you are, few ARM on remote place, no way to netinstall

BTW, does somebody know from which v6 version exactly started this strange problem ?
maybe there is a brute way to downgrade v6 below problematic version and go back from individual to bundled ?

pe1chl · Tue Mar 29, 2022 9:21 pm

It is long ago, and not a good idea to downgrade that far. You need to wait until either MikroTik fixes this problem or you have the opportunity to go on location and netinstall.
(I would not recommend upgrading a remote system to v7 anyway at this time)

Tue Mar 29, 2022 9:35 pm

(I would not recommend upgrading a remote system to v7 anyway at this time)

Depends on what you need and what has proven already to work.
Just today I upgraded 3 devices 930km away (from 7.1.1 and 7.2rc2 to 7.2rc5). First CAPAC, then CAP Lite and lastly SXT LTE (and while at it, it was also suggested to upgrade LTE firmware from V029 to V033),
No fancy configs, just simple wifi, capsman and basic routing.
Smooth sailing.

Obviously it was a calculated decision. Not this weekend but the weekend after I'm there myself :wink:

pe1chl · Tue Mar 29, 2022 9:38 pm

It is of course less risky to upgrade from 7.1.1 to another v7 version than from v6 to v7.
But even then it is necessary to do a /export before and after the upgrade to make sure no config is lost.
And in case critical config is lost, you will be in trouble.
(e.g. when you have an IPsec VPN with PSK as your only access to the router, an upgrade may lock you out because the "IPsec identity" records are often lost)

Recently I upgraded a router from 7.2rc3 to 7.2rc5 and I also lost access to it due to bugs with multiple routing tables.

mkx · Tue Mar 29, 2022 9:54 pm

(I would not recommend upgrading a remote system to v7 anyway at this time)
Depends on what you need and what has proven already to work.
...
Obviously it was a calculated decision. Not this weekend but the weekend after I'm there myself :wink:

Sure. That's why I wrote "I would not recommend upgrading" ... which is not the same as "I would advise against upgrading" or "I discourage upgrading". For a random user coming by v7 doesn't offer much that v6 doesn't (notable exception are wifiwave2 drivers on select devices and wireguard).
On the otehr hand, I always recommend upgrading to latest long-term version (6.48.6) over staying at much older versions ... notice the difference?

ksteink · Wed Mar 30, 2022 6:24 am

Well as a follow up I decided to upgrade additional spare devices that I have around:
- Another hEX S --> No issues upgrading
- RB2011 --> No issues upgrading
- CRS305 --> No issues upgrading
- hAP AC --> Same error as the CRS326

On the hAP AC I did a full reset configuration to have with the factory default configuration and tried again to upgrade with the same error as the CRS326. So the issue is not just on ARM devices.

See the Screenshot

ROSv7 Storage Error hAP ac.png

mkx · Wed Mar 30, 2022 8:08 am

The error messages are telling you that you most probably have unbundled installation of 6.49.5 on your hAP ac ... and this kind of installation can not be upgraded to monolithic v7. The last error (insufficient space) is bogus actually. You have to netinstall the device.

BrateloSlava · Wed Mar 30, 2022 8:45 am

How I upgraded my devices, which have 16 MB of internal memory, from version 6 to version 7.
For example, cAP ac, hAP ac2 and etc.

Made a backup. Text and binary.
Disabled/stopped WiFi interfaces.
Removed ALL packages from the device except these: dhcp, security and system.
Updated to version 7 in a regular way (system - package - ...)
Be sure to check, that the settings for WiFi interfaces after the update have the same parameters as in the 6th version. Interface names, bridge and VLAN parameters.
Resumed the work of WiFi interfaces.
Checked the settings.

I warn you: it is impossible to update remotely in this way.

osc86 · Wed Mar 30, 2022 11:33 am

is the ospf / instances / redistribute option supposed to work in 7.1.5? For me it's not working, even when nothing is checked, at least all connected routes are redistributed.
I solved it with route filters for now.

lvader · Wed Mar 30, 2022 1:24 pm

Small fix/suggestion for default configs:
if l2tp vpn enabled, those rules are generated. However, this will not work if IPSec enabled for l2tp, the other mikrotik l2tp-clients can't establish connections with it.

/ip firewall filter add action=accept chain=input comment="allow IPsec NAT" dst-port=4500 protocol=udp
/ip firewall filter add action=accept chain=input comment="allow IKE" dst-port=500 protocol=udp
/ip firewall filter add action=accept chain=input comment="allow l2tp" dst-port=1701 protocol=udp

Those rules should be added:

/ip firewall filter add action=accept chain=input comment="allow IPSec AH" protocol=ipsec-ah
/ip firewall filter add action=accept chain=input comment="allow IPSec ESP" protocol=ipsec-esp

PS. funny enough, those rules are generated for IPv6, but not for IPv4 for some reason...

/ipv6 firewall filter add action=accept chain=input comment="defconf: accept ipsec AH" protocol=ipsec-ah
/ipv6 firewall filter add action=accept chain=input comment="defconf: accept ipsec ESP" protocol=ipsec-esp

pe1chl · Wed Mar 30, 2022 1:32 pm

Those ESP rules are only required when L2TP/IPsec is used without NAT involved. Probably a rare thing, these days. The AH rule should not be required for L2TP/IPsec in default config.
The rule for port 1701 would better be extended with: ipsec-policy=in,ipsec

ksteink · Wed Mar 30, 2022 5:54 pm

The error messages are telling you that you most probably have unbundled installation of 6.49.5 on your hAP ac ... and this kind of installation can not be upgraded to monolithic v7. The last error (insufficient space) is bogus actually. You have to netinstall the device.

Thanks for the insight. I want to avoid Netinstall as I have several remote locations and the onsite persons doesn't have technical skills to do this remotely. I am looking for Mikrotik to provide a solution around the unbundled installation. And agreed with the bogus space error, I saw that before in v6 that I think (if I remember well) I sorted it out by going unbundled

sirbryan · Wed Mar 30, 2022 6:08 pm

is the ospf / instances / redistribute option supposed to work in 7.1.5? For me it's not working, even when nothing is checked, at least all connected routes are redistributed.
I solved it with route filters for now.

7.2rc6 has some fixes for redistribution of OSPF & RIP into OSPF via route filters.

aacable · Thu Mar 31, 2022 12:50 pm

I deployed 7.15 at few routers & had to revert back in few hours !

Either it's c2004 series, CCR1036/1016 etc,. OR the role is pppoe / gateway router | 7.15 seems to be very unstable.

Jotne · Thu Mar 31, 2022 2:50 pm

Does not look god. Memory does still be eaten up. Did try to turn of DoH som 6 hour ago, still going up. So in some days it reaches 100%.
RB 750Gr3 Router OS 7.1.5
Will send file to support.
.

Thu Mar 31, 2022 2:52 pm

Might be interesting to see what happens when it does reach 100%... completely lock up or will it start cleaning up ?

infabo · Thu Mar 31, 2022 3:43 pm

I have no issues with DOH (nextdns.io). 177mib free memory (of 256mib). uptime with current 7.1.5 1w1d17h16m37s. 7.1.3 had uptime 21+ days and no memory leak. Maybe platform-specific?

kenyloveg · Thu Mar 31, 2022 5:49 pm

Forward in dns static still doesn't work when only DoH server is enabled.

ksteink · Thu Mar 31, 2022 9:42 pm

I think I have a new bug.. Here is the situation:

- I have an RB5009 that has a S2S VPN using L2TP/IPSec to a remote hEX S (let's call it jump network server).
- The remote hEX S was using v6.49.5 but few days ago I upgrade it to version 7.1.5.
- The hEX S has L2TP/IPSec S2S VPNs to multiple clients.
- Testing 3 different remote clients (1 x RB4011 with remote IP 10.1.97.1, 1 x hAP AC2 with remote IP 10.1.64.1 and 1 x hEX S with remote IP 10.1.32.2).
- When I need to access the remote clients I go via the S2S VPN from my RB5009 into the hEX S and via noutive routing go via the S2S VPN to any of the 3 remote routes (RB4011, the other hEX S and the hAP AC2).
- The RB4011 and the hAP AC2 responds to ping but I cannot do either SSH or Winbox directly (The connection hangs like waiting to complete the 3 way handshake). The remote hEX S doesn't even respond to ping from my RB5009. but from the hEX S that has the S2S VPN I can ping it.

All my problems started when I upgraded the remote hEX S (let's call it jump network server) from 6.49.5 to 7.1.5. Before that I could access all the remote devices without a problem.

Let me add tests from RB5009:

To the RB4011: RB5009 --> hEX S (jump network server running 7.1.5) --> RB4011

[admin@MAK-IE01] > ping 10.1.97.1
SEQ HOST SIZE TTL TIME STATUS
0 10.1.97.1 56 63 177ms173us
1 10.1.97.1 56 63 180ms281us
sent=2 received=2 packet-loss=0% min-rtt=177ms173us avg-rtt=178ms727us max-rtt=180ms281us

[admin@MAK-IE01] > /tool traceroute 10.1.97.1
Columns: ADDRESS, LOSS, SENT, LAST, AVG, BEST, WORST, STD-DEV
# ADDRESS LOSS SENT LAST AVG BEST WORST STD-DEV
1 192.168.8.106 0% 4 155.1ms 158.4 155.1 160.7 2.4
2 10.1.97.1 0% 3 172.4ms 174.2 170.5 179.7 4

[admin@MAK-IE01] > /system/ssh user=admin 10.1.97.1 --> Here it hangs and never prompts for credentials

To the hAP AC2: RB5009 --> hEX S (jump network server running 7.1.5) --> hAP AC2

[admin@MAK-IE01] > ping 10.1.64.1
SEQ HOST SIZE TTL TIME STATUS
0 10.1.64.1 56 63 180ms998us
1 10.1.64.1 56 63 178ms718us
2 10.1.64.1 56 63 182ms464us
3 10.1.64.1 56 63 174ms432us
4 10.1.64.1 56 63 188ms320us
sent=5 received=5 packet-loss=0% min-rtt=174ms432us avg-rtt=180ms986us max-rtt=188ms320us

[admin@MAK-IE01] > /tool/traceroute 10.1.64.1
Columns: ADDRESS, LOSS, SENT, LAST, AVG, BEST, WORST, STD-DEV
# ADDRESS LOSS SENT LAST AVG BEST WORST STD-DEV
1 192.168.8.106 0% 4 155.3ms 164.1 155.3 174.1 7.7
2 10.1.64.1 0% 3 173.4ms 176 173.4 181.3 3.7

[admin@MAK-IE01] > /system/ssh user=admin 10.1.64.1 --> Here it hangs and never prompts for credentials

To the hEX S: RB5009 --> hEX S (jump network server running 7.1.5) --> hEX S

[admin@MAK-IE01] > ping 10.1.32.2
SEQ HOST SIZE TTL TIME STATUS
0 10.1.32.2 timeout
1 10.1.32.2 timeout
2 10.1.32.2 timeout
sent=3 received=0 packet-loss=100%

[admin@MAK-IE01] > /tool/traceroute 10.1.32.2
Columns: ADDRESS, LOSS, SENT, LAST, AVG, BEST, WORST, STD-DEV
# ADDRESS LOSS SENT LAST AVG BEST WORST STD-DEV
1 192.168.8.106 0% 1 155.4ms 155.4 155.4 155.4 0
2 100% 1 timeout
3 100% 1 timeout
4 0% 1 0ms

Let me add tests from hEX S (Jump Network Server):

To the RB4011: From hEX S (jump network server running 7.1.5) --> RB4011

[kirk@IMKS-RO01] > ping 10.1.97.1
SEQ HOST SIZE TTL TIME STATUS
0 10.1.97.1 56 64 26ms455us
1 10.1.97.1 56 64 20ms272us
2 10.1.97.1 56 64 17ms338us
sent=3 received=3 packet-loss=0% min-rtt=17ms338us avg-rtt=21ms355us max-rtt=26ms455us

[kirk@IMKS-RO01] > /tool/traceroute 10.1.97.1
Columns: ADDRESS, LOSS, SENT, LAST, AVG, BEST, WORST, STD-DEV
# ADDRESS LOSS SENT LAST AVG BEST WORST STD-DEV
1 10.1.97.1 0% 5 25.8ms 20.6 16.3 25.8 4

[kirk@IMKS-RO01] > /system/ssh user=admin 10.1.97.1
password: --> It type the password I get connection via SSH

To the hAP AC2: From hEX S (jump network server running 7.1.5) --> hAP AC2

[kirk@IMKS-RO01] > ping 10.1.64.1
SEQ HOST SIZE TTL TIME STATUS
0 10.1.64.1 56 64 26ms462us
1 10.1.64.1 56 64 17ms959us
2 10.1.64.1 56 64 23ms515us
sent=3 received=3 packet-loss=0% min-rtt=17ms959us avg-rtt=22ms645us max-rtt=26ms462us

[kirk@IMKS-RO01] > /tool/traceroute 10.1.64.1
Columns: ADDRESS, LOSS, SENT, LAST, AVG, BEST, WORST, STD-DEV
# ADDRESS LOSS SENT LAST AVG BEST WORST STD-DEV
1 10.1.64.1 0% 2 26.3ms 26.3 26.3 26.3 0

[kirk@IMKS-RO01] > /system/ssh user=admin 10.1.64.1
password: --> It type the password I get connection via SSH

To the hEX S: From hEX S (jump network server running 7.1.5) --> hEX S

[kirk@IMKS-RO01] > ping 10.1.32.2
SEQ HOST SIZE TTL TIME STATUS
0 10.1.32.2 56 64 16ms85us
1 10.1.32.2 56 64 10ms331us
2 10.1.32.2 56 64 10ms215us
sent=3 received=3 packet-loss=0% min-rtt=10ms215us avg-rtt=12ms210us max-rtt=16ms85us

[kirk@IMKS-RO01] > /tool/traceroute 10.1.32.2
Columns: ADDRESS, LOSS, SENT, LAST, AVG, BEST, WORST, STD-DEV
# ADDRESS LOSS SENT LAST AVG BEST WORST STD-DEV
1 10.1.32.2 0% 2 10.2ms 11.1 10.2 12 0.9

[kirk@IMKS-RO01] > /system/ssh user=admin 10.1.32.2
password: --> It type the password I get connection via SSH even from the RB5009 it doesn't even respond to ping.

I have checked routes in both routers and they are good:

From RB5009:

[admin@MAK-IE01] > /ip/route/print
Flags: D - DYNAMIC; X, I, A - ACTIVE; c, s, d, v, y - COPY
Columns: DST-ADDRESS, GATEWAY, DISTANCE
# DST-ADDRESS GATEWAY DISTANCE
DAv 10.1.32.0/19 <l2tp-l2tp-imks-ro01> 1
DAv 10.1.64.0/19 <l2tp-l2tp-imks-ro01> 1
DAv 10.1.96.0/21 <l2tp-l2tp-imks-ro01> 1

From the hEX S (Network Jump Server)

[kirk@IMKS-RO01] > /ip route/print
Flags: D - DYNAMIC; A - ACTIVE; c, s, d, v, y - COPY
Columns: DST-ADDRESS, GATEWAY, DISTANCE
# DST-ADDRESS GATEWAY DISTANCE
DAc 192.168.8.100/32 l2tp-mak-out1 0

Note: I do Source NAT from the RB5009 to the hEX S with the 192.168.100/32 and I have those correctly working

Some my questions are:
- Why can I ping the remote devices but cannot connect to them via SSH and Winbox?
- Why I can ping 2 routers from the RB5009 but there is one that doesn't even respond to ping.

Looks to me that there is a weird or serious bug here based on this behavior. I can access remote devices if I do an SSH from my RB5009 to the hEX S jump server and from there I do another SSH to the target router instead of going directly.

osc86 · Thu Mar 31, 2022 10:08 pm

7.2rc6 has some fixes for redistribution of OSPF & RIP into OSPF via route filters.

yes, my issue is fixed with 7.2rc7

dfdf · Fri Apr 01, 2022 12:00 am

Hello,
in 6.x.x ROS there was "unreachable" property for every route in /ip route
It's gone in 7.x.x? Why? And what should I use instead of it in scripts?

pe1chl · Fri Apr 01, 2022 12:09 am

Apparently, "unreachable" is no longer supported by the newer kernel used in RouterOS v7. You can now only use "blackhole" for a similar function.
(it is not the same, as it just drops the packets instead of returning a "destination unreachable" ICMP to the source)

eworm · Fri Apr 01, 2022 12:13 am

You can create a bridge "unreachable" without ports and addresses, then route the traffic there. Now create a firewall rule to match packets to that interface and reject.

dfdf · Fri Apr 01, 2022 12:17 am

You can now only use "blackhole" for a similar function.

I just want to filter out unreachable routes, it's absolutely unrelated to blackhole property, as blackhole is another type of route.
How can I do this in 7.x.x?
It was a fundamental property on which many scripts rely on. Why it's removed (but winbox STILL showing it properly -- as shown on my screenshot)?

dfdf · Fri Apr 01, 2022 12:19 am

You can create a bridge "unreachable" without ports and addresses, then route the traffic there. Now create a firewall rule to match packets to that interface and reject.

/ip route print. I don't want to create anything, just to filter out unreachable routes in PRINT command.

pe1chl · Fri Apr 01, 2022 12:33 am

Maybe you can work around it by using: where check-gateway and inactive

dfdf · Fri Apr 01, 2022 12:38 am

Maybe you can work around it by using: where check-gateway and inactive

check-gateway is unrelated to unreachable as it's a method to check reachability, not a result of checking.
will try to use "/ip route print where inactive and !disabled"... still not sure this will cover all possible states and is equivalent to "unreachable"

Amm0 · Fri Apr 01, 2022 5:17 am

Maybe you can work around it by using: where check-gateway and inactive

And, without "dynamic-in" route filters in V7, you can't set "check gateway" on a dhcp-client injected route in ROS V7, as you could in V6. So that approach is still limited particular setups, vs something generic for a script to find "unreachable".

BUT...

check-gateway is unrelated to unreachable as it's a method to check reachability, not a result of checking.
will try to use "/ip route print where inactive and !disabled"... still not sure this will cover all possible states and is equivalent to "unreachable"

ROS V6 has a wonderful command that just tell you the "route resolution" in

/ip route check 8.8.8.8
.       status: ok
.       interface: bridge
.       nexthop: X.X.X.X

which also worked in a script to solve all the various "unreachable" things:

:global rtcheck [/ip route check 8.8.8.8 as-value once]

I too am lacking in some scriptable method in V7 have ROS tell me definitively if a dst-address is reachable. Outside of a likely complex loop of ping interfaces=

Maybe someone has some approach?

Argosy · Fri Apr 01, 2022 10:31 am

After 5 Days I had drops in the internet connection. First I tough it's a other problem, but now after rebooting my RB750Gr3 it's stable again. Do we have a memory leak???

pe1chl · Fri Apr 01, 2022 10:43 am

Maybe you can work around it by using: where check-gateway and inactive
check-gateway is unrelated to unreachable as it's a method to check reachability, not a result of checking.
will try to use "/ip route print where inactive and !disabled"... still not sure this will cover all possible states and is equivalent to "unreachable"

Please explain you actual use case then. Your example is a route that is unreachable because it does check-gateway and the check is failing.
What other cases do you know that lead to an "unreachable" route in v7?

Fri Apr 01, 2022 10:58 am

After 5 Days I had drops in the internet connection. First I tough it's a other problem, but now after rebooting my RB750Gr3 it's stable again. Do we have a memory leak???

See 16 posts up (#170).
Jotne already noticed and provided indications it is happening on some devices.

SecCon · Fri Apr 01, 2022 11:42 am

So many issues described above.
Normally I am the kind of guy that just hit "upgrade" and close my eyes, but not with this. There is not even an "upgrade" button as such.

And I am not even running anything but a default lan with some FW entries and 3 port forwards. I guess the memory leak has me most worried.

My Mikrotik Devices:
Router RB1100AHx4 (RoS 7.1.1)
Switch CRS326-24G-2S+RM (RoS 7.1.1)

pe1chl · Fri Apr 01, 2022 11:52 am

After 5 Days I had drops in the internet connection. First I tough it's a other problem, but now after rebooting my RB750Gr3 it's stable again. Do we have a memory leak???
See 16 posts up (#170).
Jotne already noticed and provided indications it is happening on some devices.

It is unlikely to be the device, it more likely is some configuration. To solve this you would really need to make a supout file and create a ticket on help.mikrotik.com/servicedesk.

AUsquirrel · Fri Apr 01, 2022 1:58 pm

Upgrade went OK. IPV4 is ok. IPV6 network stopped working. I suspect that one of the fixes has impacted my firewall rules. The upgrade did drop an OSPF router id on one of my routers.
Re-entered the ID and it dropped straight back in.

pe1chl · Fri Apr 01, 2022 3:09 pm

Upgrade went OK. IPV4 is ok. IPV6 network stopped working. I suspect that one of the fixes has impacted my firewall rules.

Cause of that is often the incorrect configuration following an often copied advice. When both PPPoE and DHCPv6 are used and both have "add default route" checkmark enabled, IPv6 does not work on v7 while it did work on v6. You need to remove "add default route" checkmark in DHCPv6 in that case, and only leave it in the PPPoE client.
It may be that it has been fixed in 7.2rc5:

*) ipv6 - do not add duplicate dynamic prefix when static already exists;

but it is not sure if that would also prevent adding duplicate dynamic prefix when another dynamic prefix already exists.

Traveller · Fri Apr 01, 2022 4:03 pm

Version 7.1.5 really moved to long term or is it a bug?

rextended · Fri Apr 01, 2022 4:13 pm

A bug?
Inside the 7.1.5 file for mipsbe is readable: "7.1.5 (stable)"

pe1chl · Fri Apr 01, 2022 6:07 pm

Version 7.1.5 really moved to long term or is it a bug?

https://twitter.com/mikrotik_build/stat ... 5669407749

eworm · Fri Apr 01, 2022 6:12 pm

I guess that is a preparation for 7.2 in stable...
Let's wait what next week brings. 😜

SiB · Fri Apr 01, 2022 8:09 pm

I guess that is a preparation for 7.2 in stable...
Let's wait what next week brings. 😜

I remember that we buy to biggest customer new CCR who works only at new RouterOS v6.. he start journey very badly and we have stable 6.4 and next one was 6.19 and all laters. So many problems this hardware and software generate then currently with ros7 I track, read all and not buy until the STABLE will be real STABLE and not feal like experiment on my customers.
I wait for ros7.2 :)

dfdf · Fri Apr 01, 2022 11:47 pm

check-gateway is unrelated to unreachable as it's a method to check reachability, not a result of checking.
will try to use "/ip route print where inactive and !disabled"... still not sure this will cover all possible states and is equivalent to "unreachable"
Please explain you actual use case then. Your example is a route that is unreachable because it does check-gateway and the check is failing.
What other cases do you know that lead to an "unreachable" route in v7?

I use this to report one of links used for failover is down. This is useful if you have more than one upstream link to provider(s).
I still can't understand why inactive is the same as unreachable. Inactive means (for me) "usable but not currently in use" but this is definitely not the same as "unreachable" (i.e. -- can't be used).

AUsquirrel · Sat Apr 02, 2022 12:10 am

Upgrade went OK. IPV4 is ok. IPV6 network stopped working. I suspect that one of the fixes has impacted my firewall rules.
Cause of that is often the incorrect configuration following an often copied advice. When both PPPoE and DHCPv6 are used and both have "add default route" checkmark enabled, IPv6 does not work on v7 while it did work on v6. You need to remove "add default route" checkmark in DHCPv6 in that case, and only leave it in the PPPoE client.
It may be that it has been fixed in 7.2rc5:
*) ipv6 - do not add duplicate dynamic prefix when static already exists;
but it is not sure if that would also prevent adding duplicate dynamic prefix when another dynamic prefix already exists.

Thank you for your reply. Unfortunately in this case. I am not using PPPoE and my border router with DHCPv6 and IPv6 is working fine. It is the internal router and the trunk link to the border router where the problem appears to be.

pe1chl · Sat Apr 02, 2022 12:12 pm

Good luck debugging it then! With such vague stories and no config export, nobody can help you from here.

akarpas · Sat Apr 02, 2022 5:14 pm

then SD card support will be fixed for CCR1009-7G-1C-1S+ devices????????? or this problem is completely ignored?

marlab · Sat Apr 02, 2022 8:40 pm

Excessive CPU usage on the RB3011 is still not fixed.
Does anyone know if Mikrotik has at least recognized the problem at some point?

I've reported the problem with regular reboots on RB3011 over a week ago, and they even didn't bother to reply. I doubt they any longer care about that router (since RB4011 is out) :/

Brendon · Sun Apr 03, 2022 1:41 pm

Hi,

after upgrade hAP ac from 6.49.4 to 7.1.5, my Realtek USB to ethernet adapter is recognized as LTE type device and not working. Same behaviour in 7.2rc7.

Vendor ID: 0x0bda
Device ID: 0x8153

mkx · Sun Apr 03, 2022 2:02 pm

That's weird indeed. Official (if anything like that exists in linux) USB IDs database lists this device as "RTL8153 Gigabit Ethernet Adapter". What does /system/resource/usb/print detail say about it?

Brendon · Sun Apr 03, 2022 3:04 pm

Hi mkx,
it is adapter with built-in usb hub, perfectly working on ros 6.xx.

Columns: DEVICE, VENDOR, NAME, SPEED
# DEVICE  VENDOR                   NAME                     SPEED
0 1-0     Linux 5.6.3 ehci_hcd     RB400 EHCI                 480
1 1-1     VIA Labs, Inc.           USB2.0 Hub                 480
2 1-1.4   Realtek                  USB 10/100/1000 LAN        480

mkx · Sun Apr 03, 2022 4:13 pm

I suggest you to open a support ticket. Also create a supout.rif file which should help MT devs to debug this problem.

bbartlomiej · Mon Apr 04, 2022 9:43 am

Upgrade from 7.1.3 to 7.1.5 ended up in bootlop for me. It's the 1st time such thing happened to me.

2x RouterBOARD 750G r3 hEX ended up in bootloop (rescue bootloader wouldn't work fully either) and had to be netinstalled. CAP ACs were OK.
1st reboot was fine, it's the 2nd one after upgrading firmware that caused it. Before 2nd reboot my ip route rule was inactive.

SiB · Mon Apr 04, 2022 10:30 am

I have 2 users with LHGGR with boot-loop - just that ros7 in general cannot be called stable or safe. Waiting for 7.2 with no harm comments.