Any details about tile improvement?!) tile - improved overall system performance and stability ("/system routerboard upgrade" required);
This one:What fix? Please don't hope that everybody knowsLooks like the ipsec fix did not make it into the release. *sigh* Do I have to wait for version 6.42?
Thanks anyway!
*) ipsec - properly detect interface for "mode-config" client IP address assignment;
I've asked this before on the RC forum, however I've had no response.Hello,
Any details about tile improvement?!) tile - improved overall system performance and stability ("/system routerboard upgrade" required);
Regards,
Bit of a long shot, however will this improve things for BGP?We have improved handling of operations on TILE based devices (CCRs). This fix will not affect anyone negatively, but might help in rare cases with different random problems. For example, there were users who used CCR1072 and CPU core was stuck on 100% usage due to IPsec HW encryption. This fix resolves it. Basically, CPU is now better handling assigned processes. There is no need to upgrade, if you do not have any problems with your device.
Yes! I have 3 CRS125's one's a desktop version other two are rack mount and I see this also...Out of sudden I found nice dialog with drop-down box named "SIM", under System -> Routerboard -> SIM:
Something that I neved expected from my old (but good) CRS125-24G-1S (https://mikrotik.com/product/CRS125-24G-1S-IN, which is essentially a 24-ports switch with decent router (like 951's) built-in, and it doesn't feature any SIM slot for what I know. Or there is a hidden one?
On my image you can see ROS and f/w versions (6.41.3), device model and button I talk about.
What's that? Hidden feature or another bug-not-a-feature thing?
I had issues with the DHCP server not handing out addresses after some time of proper functioning in 6.41.2 if the interface is a bridge. This was reported by some users, and seemed it was addressed in the RC, so I assumed a ticket would not be necessary anymore (In my case I reverted my main router to 6.40.6 which solved the issue, so no emergency on this side).docmarius - Are you referring to one of the fixes in 6.42rc version or some other problem which you have resolved/reported to support?
zoombie - Please provide more information. Do you have any problems with real life traffic not with the BTest? Can you downgrade to 6.41.2, make sure that problem is gone, upgrade back to 6.41.3 and see that problem appears once more?
zoombie - Please provide more information. Do you have any problems with real life traffic not with the BTest? Can you downgrade to 6.41.2, make sure that problem is gone, upgrade back to 6.41.3 and see that problem appears once more?
Will this help with SNMP being limited to a single core? (clipping that core to 100% if you're pooling too many parameters)?We have improved handling of operations on TILE based devices (CCRs). This fix will not affect anyone negatively, but might help in rare cases with different random problems. For example, there were users who used CCR1072 and CPU core was stuck on 100% usage due to IPsec HW encryption. This fix resolves it. Basically, CPU is now better handling assigned processes. There is no need to upgrade, if you do not have any problems with your device.
I ask since it was my picture and question initially. What will happen if I set this setting to different value ;) and, more seriously, since sim slot is not built in, does it mean firmware reports it is there by mistake and so f/w may be considered buggy?105547111 - You see this option on your device by mistake. Of course, since you do not have a SIM slot, you should not see SIM related options on CRS125 device;
poizzon - If you have a RB2011 model which has serial console port, then please provide serial output so we can see why router can not boot any more;
strods - I also see SIM option on RB433AH and on hAP ac lite. Winbox 3.12105547111 - You see this option on your device by mistake. Of course, since you do not have a SIM slot, you should not see SIM related options on CRS125 device;
We will eventually improve it, so that you only need one reboot. For now, you are free to ignore the upgrade and reboot when you need to.Is it now really necessary to update routerboard firmware everytime we update ROS since the version numbering now follows ROS version number? I cannot believe there are changes everytime and it is quite annoying to have to reboot twice.
I also have the SIM menu on RB951G-2HnD now.
I had to re-enter one of my L2pt/IPSEC connections and noticed again that it went not going to sleep despite I was on 6.40.6. So I looked in the export file on my name for that connection and saw the Neighbour Discover was enabled for that connection. This discovery gave regular a kick to the L2pt/IPSEC connection so it could not catch sleep and stayed awake all the time.update: I noticed running with the bridge the dail-on-demand pppoe was never going to sleep and I had to force an time-out to have change state to "waiting for pakets". Back on 6.40.5 it sleeps again till needed. The pppoe connection was no part of the bridge.
No X86 packages yet?
The requested URL /routeros/6.41.3/routeros-x86-6.41.3.npk was not found on this server.
https://download.mikrotik.com/routeros/6.41.3/routeros-mipsbe-6.41.3.npk
...
https://download2.mikrotik.com/routeros/6.41.3/routeros-mipsbe-6.41.3.npk
I haven't seen any information here on forum but hopefully this version patches the malware attack "Slingshot". It was only announced 2 days ago, but possibly it was known already to Mikrotik and fixed, as mentioned at the end of the article in the link below.
https://www.engadget.com/2018/03/11/sop ... h-routers/
941-2nD bricked as well.....bumpt - rb2011 - bricked
How did you resolve? I'm dealing with a router doing that right now..Hi..installed update v6.41.3 on to my RB2011UiAS-RM and now its caught in a booting loop. any ideas?
-- ISSUE RESOLVED! Thanks
That's going to be difficult if the router bricks (like mine just did).. I don't keep supout files laying around, I'd have to generate a new one every time I did any modification to my settings for it to be relevant anyhow.3) Send supout file from the old version (so we can see configuration and installed packages);
It should be done by the firmware, isn't it?Please note that upgrade process happens on the old version - not the one which you install on your router.
I would like to know the same thing.Bit of a long shot, however will this improve things for BGP?We have improved handling of operations on TILE based devices (CCRs). This fix will not affect anyone negatively, but might help in rare cases with different random problems. For example, there were users who used CCR1072 and CPU core was stuck on 100% usage due to IPsec HW encryption. This fix resolves it. Basically, CPU is now better handling assigned processes. There is no need to upgrade, if you do not have any problems with your device.
Thankstonymobile - Such features will never be added into 6.41.x version. v6.41, v6.42, v6.43, etc. will have new features. v.6.41.x, v.6.42.x, v.6.43.x, etc. will include only fixes. That is the whole point of bug fixes only versions.
So to say, the only safe option is to stay with bugfix branch and read forum even for its bugs and features. Current is too risky.This is not a version related problem and such posts might discourage others from upgrade since others might assume that this version in some way makes wireless work worse than before.
By the way, may I ask for an advice: what if a person installed full set of ROS packages (for reason unknown) - the ones the listed as "Extra packages", is there a way to have router be back on standart set of packages (named "Main package") beside deleting it one by one? Once I tried to mark as deleted all packages and upload Main package to flash and reboot and recieved bricked device and some hours of netinstall.Upgrade tests are made for each RouterOS release and version is released when upgrade works perfectly on all the lab units. Upgrades usually fail for end users due to:
1) Power issues;
2) Damaged storage;
3) Full storage;
4) Wrong package installations usually caused by installed rc version in the past.
For example, upgrade with bundle package should never fail.
We have re-worked upgrade in 6.42 version in order to avoid #3 and #4 caused problems as much as possible.
What I see on this topic the process of ROS update/downgrade should be described in details?However, answer is simple - you can downgrade or upgrade device away from the version which you use by simply dropping bundle package into files section and either upgrading or downgrading the device.
Hi strods,...
mpadmin - About what kind of "bug" are you talking about? Have you contacted with support and figured out that there is an actual software issue with DHCP on your network?
...
And your answer was:I had issues with the DHCP server not handing out addresses after some time of proper functioning in 6.41.2 if the interface is a bridge. This was reported by some users, and seemed it was addressed in the RC, so I assumed a ticket would not be necessary anymore (In my case I reverted my main router to 6.40.6 which solved the issue, so no emergency on this side).
So I plan to downgrade all my routers, because bridge and dhcp server on it is a must (small routers with WiFi). Is there any workaround for the dhcp server hang on bridge before the 6.42 release?...
docmarius - This problem required re-work of DHCP service functionality on bridge interface. Since it was more like a re-written functionality not a fix, we decided to leave this for 6.42 release and not include in 6.41.x for safety reasons;
...
Thanks for pointing out. I was not using large data transfers, so I would find out in a month or so, but as I see it, standard TX Rate was showing between 100-300 Mbit on my network interface, now I see only 40Mbit...Hi,
A have a new one HAP AC^2. (D52G-5HacD2HnD-TC)
The 2.4GHz network have better bandwith then 5.0GHz. Does anyone have the same? ?
Best Redgards,
I didn’t have this problem in 6.41.2, last time I checked, before upgrade it was still on the speeds I have mentioned in previous post. I will contact the support also.pawelkopec88 - If you had already this problem in 6.41.2, then please in the future at least mention this in forum or contact support team first. This is not a version related problem and such posts might discourage others from upgrade since others might assume that this version in some way makes wireless work worse than before. Please contact support@mikrotik.com and try to debug your wireless issue through an e-mail.
All affected customers reached, unplugging power waiting 10 seconds and then plugging power back again made them connect to basestations again.Hello Folks!
Upgrade successful on rb411, rb333, sxt, sextant, qrt5, rb711, crs.
Sadly I lost all remote Dynadish devices, after upgrading they never got contact with the wireless base stations again.
I have no ability to visit customers at this time, so I can not tell if there is any ethernet connection to it or what happened.
Dynadish was okey before upgrade, and had 6.37.5 before upgrade.
I asked them to power off and on the Dynadishes when they come home after work, lets see if that works or I have to travel some 300Km to fix it, annoying.
I think this is what those people, who sometimes accuse MikroTik of not being transparent enough, are talking about. Sure, "smb - improved NetBIOS name handling and stability" is technically correct, fixing buffer overflow does improve stability. But exploitable buffer overflow sounds like quite important detail that should be mentioned somewhere on MikroTik's site.Another bug fixed in this version: https://www.coresecurity.com/advisories ... r-overflow
...and should be incorporated into the bugfix branch, for those of us who want to patch the vulnerability, without also having to rework our complex bridge/VLAN/DHCP configurations to work with the new implementation in 6.41.I think this is what those people, who sometimes accuse MikroTik of not being transparent enough, are talking about. Sure, "smb - improved NetBIOS name handling and stability" is technically correct, fixing buffer overflow does improve stability. But exploitable buffer overflow sounds like quite important detail that should be mentioned somewhere on MikroTik's site.Another bug fixed in this version: https://www.coresecurity.com/advisories ... r-overflow
zoombie - To which IP address does your router resolve upgrade.mikrotik.com?
.dhcp_server offering lease 192.168.88.X for XX:XX:XX:XX:XX:XX without success
It can't, because when you want to route connection to another WAN, you need to start with very first SYN packet. But the info needed by tls-host only comes later, and then it's too late to route connection elsewhere.tls-host does not work with "mark-routing" mangle rules.
You could match tls-host and create a dynamic address-list entry for the address. Then match new connections for these addresses for routing-mark. You miss the first connection, but it is worth a try.It can't, because when you want to route connection to another WAN, you need to start with very first SYN packet. But the info needed by tls-host only comes later, and then it's too late to route connection elsewhere.tls-host does not work with "mark-routing" mangle rules.
Adventurous? Should be pretty straight forward. Something like this should work:Not feeling that adventurous. Too bad Google uses the same IP blocks for everything; otherwise, I could have simply created a list for YouTube and used that.
/ip firewall filter
add action=add-dst-to-address-list chain=forward address-list=youtube protocol=tcp tls-host=*.youtube.com
/ip firewall mangle
add action=mark-routing chain=prerouting connection-state=new dst-address-list=youtube new-routing-mark=youtube
routeros-mipsbe-6.41.3.npk contains almost all the packages you have installed separately:Hi,
just bricked two of my three RB962UiGS-5HacT2HnT by upgrading from 6.41.2 to 6.41.3.
I did the following steps - as I did with the last update:
1. uploaded routeros-mipsbe-6.41.3.npk
2. reboot
3. uploaded active modules
- advanced-tools-6.41.3-mipsbe.npk
- dhcp-6.41.3-mipsbe.npk
- ipv6-6.41.3-mipsbe.npk
- multicast-6.41.3-mipsbe.npk
- ntp-6.41.3-mipsbe.npk
- routing-6.41.3-mipsbe.npk
- security-6.41.3-mipsbe.npk
- system-6.41.3-mipsbe.npk
- user-manager-6.41.3-mipsbe.npk
- wireless-6.41.3-mipsbe.npk
4. /system routeros upgrade
5. reboot
and now the system is in an endless reboot loop where it beeps alternating at a loud beep and less loud afterwards.
Whats wrong and how could I get them back to life?
Zap.
Enabling the interface again after dismissing the warning, enables the interface, however this does create an issue for me as I use the scheduler to switch pppoe accounts, the above error causes the account not to activate at the correct time.Couldn't change Interface <TelkomInternet> - object doesn't exist (4)
To keep rolling until the actual issue gets fixed, check whether use of a standaloneEnabling the interface again after dismissing the warning, enables the interface, however this does create an issue for me as I use the scheduler to switch pppoe accounts, the above error causes the account not to activate at the correct time.
:do {...}
on-error {...}
{
:local attempts 3
:while ([/interface pppoe-client print count-only where disabled=no && name=pppoe-out1]=0 and $attempts>0) do={
:do={/interface pppoe-client enable [find name=pppoe-out1]};
:set attempts ($attempts-1)
}
}
execute script=" ... "
do={ ... }
I've encountered the same problem. Details in post #80 above.Hey MikroTik Team,
I'm getting a strange issue, upon re-enabling my PPPoE-client, Winbox spits out the following error:
...Couldn't change Interface <TelkomInternet> - object doesn't exist (4)
I have the same with an HP Printer connected over wifi through a non MT device. This has been so with every MT firmware. I believe the device requesting the IP is to blame.Problem with DHCP Server in this (6.41.3) release!
I periodically see series warning messages from some clients in Log. Mesages have the form:.dhcp_server offering lease 192.168.88.X for XX:XX:XX:XX:XX:XX without success
Thanks for the heads up, thought I was the only one after not finding anything, hence I had to register just to report thisI've encountered the same problem. Details in post #80 above.
viewtopic.php?p=650142&sid=12116311e944 ... 85#p648907
Duly noted. BUT is it just an increase and nothing to bother with or is it out of MT SPEC for MAXWattage witch is what you should calculate in your DC or TelCo Rack not to hit stuff down the road.Word of !WARNING for anyone who has the CCR1072-1G-8S+.
We have two of these units, since the upgrade both have used consistently 10 more watts of power!
This has also increased the temperature of the device and fan speed, that can't be a good thing can it?
We've contacted Mikrotik and this is their response:
***
Power consumption is caused by this improvement:
*) tile - improved overall system performance and stability ("/system routerboard upgrade" required);
This fix will make your CCR work more stable.
***
Just be aware if you're in a data centre environment with limited power.
We have also upgraded our two CCR1036-8G-2S+ which only saw a slight increase in power usage after the upgrade.
Before the update they were both consuming around 45W each, now they are consuming around 55W each.Duly noted. BUT is it just an increase and nothing to bother with or is it out of MT SPEC for MAXWattage witch is what you should calculate in your DC or TelCo Rack not to hit stuff down the road.Word of !WARNING for anyone who has the CCR1072-1G-8S+.
We have two of these units, since the upgrade both have used consistently 10 more watts of power!
This has also increased the temperature of the device and fan speed, that can't be a good thing can it?
We've contacted Mikrotik and this is their response:
***
Power consumption is caused by this improvement:
*) tile - improved overall system performance and stability ("/system routerboard upgrade" required);
This fix will make your CCR work more stable.
***
Just be aware if you're in a data centre environment with limited power.
We have also upgraded our two CCR1036-8G-2S+ which only saw a slight increase in power usage after the upgrade.
Direct Question then is it out of the SPEC or not?
Faced this with RBLHGG-5acD, fixed 6.42rc52.Hello.
I have a problem with RBLHGG-5acD. I upgrade the software and firmware up to 6.41.3.
When I change some settings in "wireless" like "scan list" or "wireless protocol", the device reboots.
After this settings are not changed and I see in log "router was rebooted without shutdown" "Kernel failure in previous boot"
This is the first time I've come across this.
Help me please.
hAP ac lite does show the switch menu in WebFig, hAP ac² doesn't, while both of them run the same version (6.42rc49 at the moment of checking this). I cannot believe Mikrotik would remove the switch menu before having the complete functionality available via bridge configuration, so does this really mean that for ar8327, VLAN filtering configuration of bridge already completely manages the VLAN filtering in the switch chip?Switch menu was removed recently because the hw accelerated bridges are the way mikrotik goes these days.
/interface ethernet switch
It's clear, but it's bad.Switch menu was removed recently because the hw accelerated bridges are the way mikrotik goes these days.
Where can I see "Overall stats", "Rx Stats" and "Tx Stats" for wired interfaces?
What makes you think it is a bug ? If you play with the interface you are connected with, you can expect that behavior.Anybody experiencing the extremely frustrating bug whereby you go to add an interface to a bridge and it just disconnects you from Winbox??
I went a bit deeper into it and setting VLAN filtering at bridge level just disables hardware acceleration (without MSTP so that's not the reason) but has no visible effect inhAP ac lite does show the switch menu in WebFig, hAP ac² doesn't, while both of them run the same version (6.42rc49 at the moment of checking this). I cannot believe Mikrotik would remove the switch menu before having the complete functionality available via bridge configuration, so does this really mean that for ar8327, VLAN filtering configuration of bridge already completely manages the VLAN filtering in the switch chip?Switch menu was removed recently because the hw accelerated bridges are the way mikrotik goes these days.
The fact that the wholetree is still there in the CLI cools my optimism down a bit, though.Code: Select all/interface ethernet switch
/interface ethernet switch port
/interface ethernet switch vlan
IsAnd what about this (hAP ac²)?Where can I see "Overall stats", "Rx Stats" and "Tx Stats" for wired interfaces?
/interface ethernet print stats
/interface ethernet switch print stats
From console is ok, but what about show this in winbox & web-iface?IsAnd what about this (hAP ac²)?Where can I see "Overall stats", "Rx Stats" and "Tx Stats" for wired interfaces?andCode: Select all/interface ethernet print stats
not what you want?Code: Select all/interface ethernet switch print stats
Hello HolgerHi Boris,
I have this kind of problem most times, especially when transfering a configuration to another kind of device. Try this:
If you have now idea how to correct, login to a working system on target version (e.g. install "old" version with your config and do normal upgrade with in system/packages menu) and use the export subcommand (e.g. problem in bridge config -> /interface bridge export). You also can do the defective config over the menus and show the result with the export subcommand.
- Have a test router in your local network - upgrade it.
- Reset configuration without using a configuration script (system reset-configuration no-defaults=yes)
- Open your script file in editor
- Connect to the test router with winbox e.g. via mac address and open "new terminal"
- Mark some lines from script file and use copy paste to terminal in winbox (winbox paste in right mouse menu)
- If you get an error look at the line and try to correct it
Holger
I have 6.38.3 version, and I want to upgrade, but I want to avoid DHCP Server fail on bridge. What version is afected? Can I upgrade to 6.40.6 without risk of this bug?105547111 - You see this option on your device by mistake. Of course, since you do not have a SIM slot, you should not see SIM related options on CRS125 device;
docmarius - This problem required re-work of DHCP service functionality on bridge interface. Since it was more like a re-written functionality not a fix, we decided to leave this for 6.42 release and not include in 6.41.x for safety reasons;
zoombie - Why can not you upgrade? Upgrade functionality is not working or simply you did decide to stick with 6.40.3 version? I recommend that you contact support@mikrotik.com. Sounds like conversion from old bridge implementation to new one might have affected you, but we can not comment on this further without debugging your specific case through support;
guipoletto - This feature does not make single core processes multi-core, this does not make CPU more powerful. This new fix can help you in cases if device works abnormally and, for example, gets stuck on single core 100% usage which could lead up to Watchdog reboots.
IMHO yes, but better to ask here >> viewtopic.php?f=21&t=131129&start=50I have 6.38.3 version, and I want to upgrade, but I want to avoid DHCP Server fail on bridge. What version is afected? Can I upgrade to 6.40.6 without risk of this bug?
Hello Boris,Hello Holger
Thank you for your advice, but.....
I did like you sugested and find a lot of changes in syntax. I don't understand why but this is Mikrotik developer decision. I made changes in my .rsc file and still doesn't work so I start to configure from scratch (on latest firmware version).
I finish with configuration and Now I expect that export and than import should work. I export file , download this file, I take another router, upload file and run command in terminal:
[admin@MikroTik] > system reset-configuration no-defaults=yes run-after-reset=config.rsc
Some parts of the configuration are OK, other are not. I also try through Winbox with Reset configuration , x- No default Configuration and than I select from dropdown menu config.rsc file.
After configuratin was reset I see that IP address on LAN in 0.0.0.0/0 ......
Now I realy need some help, because I would not like to configre 50pcs (now and more in the future) manualy
RB2011UAS-2HnD ROS v6.41.3
Enabling pppoe client interface, waiting for a number of seconds (>14 secs), then disabling it results in "no such item (4)" error.
Occurs in Winbox and Terminal.
Also happens in 6.42rc43
[admin@teszt] /interface> set ppp disabled=yes;
no such item (4)
[admin@teszt] /interface> set ppp disabled=yes;
[admin@teszt] /interface> /delay 20; set ppp disabled=no
no such item (4)
[admin@teszt] /interface> set ppp disabled=no
[admin@teszt] /interface>
Looks like you need some network engineer for basic troubleshooting of the problem you have after update. Like ping, traceroute...So yes. We need a fix for this.
NEED HELP! ASAP!!!!!
Looks like you need some network engineer for basic troubleshooting of the problem you have after update. Like ping, traceroute...So yes. We need a fix for this.
NEED HELP! ASAP!!!!!
We ran into a similar problem when upgrading from 6.40.6 to 6.41.3. It may be due to the re-implementation of bridges in 6.41, which seems to result in some combinations of features working differently (i.e., not as intended) under the new version. In our case, I think the change in behavior pertains to our use of VLANs. (As soon as I get the time to build a testbed, and figure out what we need to change in our standard configuration to make it work the same in 6.41.x, I'll post the findings.)Nope.. I can ping out to google dns, yahoo without issues. I can traceroute without issues. I have access to all my equipment behind my Mikrotik without issues.
So it's a firmware issue that needs to get fixed ASAP. Shit isn't rocking science kid.
Obviously, if it works fine on all firmware below 6.38.7 then it's something with that firmware not my configs or my networks
Yeah, but we still don't know anything about the issue except that "Internet is not working". For example, what error does your browser show when you're trying to open some page? Did you use Torch/Firewall rules/Packet sniffer to see where exactly packets are lost?Nope.. I can ping out to google dns, yahoo without issues. I can traceroute without issues. I have access to all my equipment behind my Mikrotik without issues.
So it's a firmware issue that needs to get fixed ASAP. Shit isn't rocking science kid.
Yeah, but we still don't know anything about the issue except that "Internet is not working". For example, what error does your browser show when you're trying to open some page? Did you use Torch/Firewall rules/Packet sniffer to see where exactly packets are lost?Nope.. I can ping out to google dns, yahoo without issues. I can traceroute without issues. I have access to all my equipment behind my Mikrotik without issues.
So it's a firmware issue that needs to get fixed ASAP. Shit isn't rocking science kid.
We ran into a similar problem when upgrading from 6.40.6 to 6.41.3. It may be due to the re-implementation of bridges in 6.41, which seems to result in some combinations of features working differently (i.e., not as intended) under the new version. In our case, I think the change in behavior pertains to our use of VLANs. (As soon as I get the time to build a testbed, and figure out what we need to change in our standard configuration to make it work the same in 6.41.x, I'll post the findings.)Nope.. I can ping out to google dns, yahoo without issues. I can traceroute without issues. I have access to all my equipment behind my Mikrotik without issues.
So it's a firmware issue that needs to get fixed ASAP. Shit isn't rocking science kid.
Obviously, if it works fine on all firmware below 6.38.7 then it's something with that firmware not my configs or my networks
In any case, I'd try upgrading this one location to 6.40.6. If that works, then you may be running into the same problem that we are.
FWIW, our standard configuration in a Multi-Unit Dwelling (MUD) works something like this:While different. This sounds similar to what I'm experiencing. Router A cannot ping router F, but I can ping router F using router A as a gateway. Router A can establish a vpls tunnel to router F. There are some oddities that didn't present themself in 6.38.5 for me. I changed some path costs and made things stable, but if anything in the network changes it can resort to previous issues.
Yours appears to be in reverse where you're endpoints can't use them as a gateway.
I can make it do what you're doing, where I can't use it as a gateway but router A can ping F. But using A as my gateway I cannot ping F.
[admin@SXT_ROUTER_xxx] > interface pppoe-client set disabled=no 0
no such item (4)
[admin@SXT_ROUTER_xxx] > interface pppoe-client set disabled=no 0
[admin@SXT_ROUTER_xxx] >