Hello,
I would like to describe big problem with RB2011. I have a lot of that routerboards in my network and I use always the same topology:

ehter1 is unused (since it is service port)
ether2 is uplink (here is usually used an Groove as client of an P2P link)
ether3 - ether9 are used for Grooves or other devices cofigured in Access Point mode (for clients and P2P links) all that ports are bridged together.

Once in a while RB2011 freeze on gigabit ports. When it happens, all ports ether1-ether5 are switched together (when it works normaly are bridged only ports ether3-ether5) and I can not ping/find Mikrotik on that gigabit ports. But when I use potrs ether6-ether10, I can see Mikrotik normaly.

I tryed bulitin switches insted of bridge (so port ether3 as master, ether4 and ether5 slave, also switched ports ether6-ether9 and ports 3 and 6 bridged together) but without any effect.

I tryed many versions of RouterOS - etc 5.24, 5.23, 6rc14 - still the same.

Always helped powercycle or reboot throught fastehternet ports.

Do you have the same problem as me or an solution?

Yes, we have the same problem

I sended report of that problem to mikrotik support but currently withou any answer (well, I got only automated reply).

I will give you report when I will see any progress.

Regards
Pelish

Did you both try v6.1 where these issues should be fixed?
http://www.mikrotik.com/download/share/ ... be-6.1.npk

Thank you for reply. I am currently trying that version. I will give report when in still occurs.

Did you both try v6.1 where these issues should be fixed?
http://www.mikrotik.com/download/share/ ... be-6.1.npk

Is this solved in 5.25?

I have no problem

I have no problem

I've some RB2011s without problems and one with this problem. Problem is once a month.

Definitely it is not a problem of all RB2011s or better - it sometimes occurs on some places. I still have many RBs running for months without any problem. Also it seems that when I use fastethernet ports instad of gigabit ports it is not so often.

Hopefully 6.1 fixing that so we will see...

So RouterOS 6.1 didn't fix that. I am sending suppouts to support once again.

Regards
Pelish

So RouterOS 6.1 didn't fix that. I am sending suppouts to support once again.

Regards
Pelish

Please make sure to also upgrade your RouterBOOT version with this command "/system routerboard upgrade". Let us know how it works after that.

Yes I have upgraded RouterBOOT. I do that every time when I am upgrading my boards.

I'm seeing the same thing. One Ethernet plugged into eth1, one into eth9, one into eth10. eth1 as lan port, eth10 as wan port, eth9 as another wan port. DHCP server, masquerade, uPnP, PPPoE. Really simple stuff.

Unplugging eth1 and plugging it back in resets the switch chip or something, and it starts working again, no problem.


uptime: 4d23h3m30s
version: 6.1
build-time: Jun/12/2013 11:50:54
free-memory: 31.3MiB
total-memory: 64.0MiB
cpu: MIPS 74Kc V4.12
cpu-count: 1
cpu-frequency: 600MHz
cpu-load: 20%
free-hdd-space: 111.1MiB
total-hdd-space: 127.8MiB
write-sect-since-reboot: 2626
write-sect-total: 32077
bad-blocks: 0%
architecture-name: mipsbe
board-name: RB2011LS



routerboard: yes
model: 2011LS
serial-number: 3D7102D7DDA0
current-firmware: 3.08
upgrade-firmware: 3.08

normis,

Could you reply to my previous post about seeing 2 different hardware versions of RB-2011UAS-RM?

I believe this could be one of the variables here.

Example:

RB-2011UAS-RM (First hardware version, stupid power plug inside) - 5.25 no problems
RB-2011UAS-RM (Second hardware version, power plug now outside) - Can't run 5.25. Completely unstable. Move to 6.1 and it works fine.
(I have not tested with 6.1 in production as it still has too many bugs to be considered a useful option for us)

My question is simple: Are there 2 different hardware versions of RB-2011xxxx boards?
This is probably what is causing some to be stable with X firmware, and some to not be stable with the same.

I can confirm that after 188 days of uptime, my RB 2011 have the same simptoms. The gigabit ports stop working and the router is not accesible until restart.

Similar problems here with RB2011L. Router just reboots randomly at 5.x. Once upgraded to 6.1 the router doesn't reboot anymore but VPN has too many problems for production implementation. It's always like this....two steps forward, one step back with every major upgrade.... but let the users figure out for themselves what is the problem and report back to Mikrotik. Not cool.

After upgrading to 6.1 the router worked for about one hour only. The problem remains.

After upgrading to 6.1 the router worked for about one hour only. The problem remains.

Have you upgraded routerBOOT?

several older RB2011LS-IN routers deployed with no issues.
Have recently purchased and deployed a few RB2011L-IN routers and these routers are doing just that.
They will route traffic then after a several hours to days the Gigabit Switch group seems to disconnect from the CPU and begins switching all ports in the group.
Is it a heat issue??
I will be pulling several out of service if this is not resolved, as it has adversely affected our network integrity.

I upgraded from 5.24 to 6.1 and RB freezed... Conf is not compatible with the 6.1 version... Never upgrade form 5.24 to 6.1

Any idea how this problem will be solved?

Thank you

Oh dear lawd I hope we get to the root of this problem . I have deployed so many of these guys and have had experienced the same symptoms recently.Will have to go back to 750G's + Fiber/Ethernet converters again

Any update??? Will be this problem solved on 5.X version?

Thank you

Could it be the MAC assignment on the bridge if the auo-mac option is enabled.
As far as I have seen, the MAC address of the bridge will get the MAC of the lowest ether port connected. (and obviously configured as member of the bridge).
I have 2 RB2011UAS-2HnD-IN units connected with a VLAN trunk. I can do a little testing within a couple of days to see what happens when ether1 get dis/connected.

I have the same problem with two 2011's. Will run for a period of time than lock up. Only a reboot fixes.

I have the same problem with two 2011's. Will run for a period of time than lock up. Only a reboot fixes.

Have you instaled ROS 6.1 ?

A customer had the same issue. We replaced the motherboard and I have no longer access to it and can't really test this out, so I'm writing this so that hopefully others may find it handy information.

I read in the changelog that this issue (or something very similar) has been fixed in a newer routerboot. The suggested rooterboot upgrade version in winbox was 2.41 - which is OLDER than the one running! How can upgrade to a newer routerboot be performed when winbox's "upgrade" is a downgrade? (Installing ros 5.24 (newest at the time of testing) did not affect this.)

We are seeing the same issues with the 2011 new routers we have. All ports go into switch mode and router is not visible via winbox or IP neighbors from adjacent routers.

We are upgrading to ROS 6.2 now but has anyone been able to confirm whether this has fixed the issue or not?

Thanks
Scott

Add me to the list, I was using 5.24 where the freeze up happened once a week. Now I am using 5.25 and the freeze up would happen every hour or less. It does seem that heavy traffic (30-50Mbps) triggers the freeze up. Hopefully someone comes up with a fix.

please upgrade to RouterOS v6.3 as there is a fix that will fix your problems:
*) RB2011L - fixed occasional gigabit switch-chip lockup;

please upgrade to RouterOS v6.3 as there is a fix that will fix your problems:
*) RB2011L - fixed occasional gigabit switch-chip lockup;

Too late, sold both units already.

If the RouterOS v6.3 didn't help we suggest to try out this RouterBOOT firmware release:
http://www.mikrotik.com/download/share/ar9344_3_09.fwf

NOTE: after upgrading the RouterBOOT to v3.09 you can't downgrade the RouterOS to v5.x

Also I couldn't get RB2011L to get past 310mbit/s throughput on giga ports. CPU 100%. Last tested on 6.2.

Also I couldn't get RB2011L to get past 310mbit/s throughput on giga ports. CPU 100%. Last tested on 6.2.

How did you test that? Was your traffic generator powerful enough?

Also I couldn't get RB2011L to get past 310mbit/s throughput on giga ports. CPU 100%. Last tested on 6.2.

Keep in mind a cable has two ends so there's always two devices involved. I can attest to your figures judging from a simple test RB2011UAS <-> RB2011UAS using ROS 6.3 traffic server and client. But what is it I measure when the CPU is maxed out? It is not how much traffic the switch chip can handle or routing speed or whatever? No. This is simply testing how much traffic the system can generate, examine and count using this specific software because it's clearl the CPU is the bottleneck here.

The art of benchmarking is knowing what to benchmark, how to do it and putting the results into perspective.

Maybe I didn't use the right words (throughput). What I mean is when I connect e.g. two Intel I7 computers with Intel E1000 cards with each other through giga ports on RB2011 and start transfering files from one to another, the CPU on RB2011L reaches 100% and max transfer speed is 310 mbit/s.

Maybe I didn't use the right words (throughput). What I mean is when I connect e.g. two Intel I7 computers with Intel E1000 cards with each other through giga ports on RB2011 and start transfering files from one to another, the CPU on RB2011L reaches 100% and max transfer speed is 310 mbit/s.

This still doesn't explain the test methodology. If you just used FTP to send file, it will not show true maximum. If your device reached 100% CPU, it means it was processing the traffic, so you must have some firewall or something.

Please tell us these things:

- what program was tranferring the data
- what data was transferred?
- how many connections
- TCP or UDP
- What configuation do you have on the router (paste output of "/export compact")

Which are the results after applying v6.3 fix? Has the freeze issue been fully resolved?

please upgrade to RouterOS v6.3 as there is a fix that will fix your problems:
*) RB2011L - fixed occasional gigabit switch-chip lockup;

Is there a way to get log or debug info for this issue when it occurs. I have a router that I think is having this problem and have just upgraded to v6.4.

Rather than just sit and wait to see if it re occurs again I would like to try and proactively monitor for the issue.

Thanks.

Maybe I didn't use the right words (throughput). What I mean is when I connect e.g. two Intel I7 computers with Intel E1000 cards with each other through giga ports on RB2011 and start transfering files from one to another, the CPU on RB2011L reaches 100% and max transfer speed is 310 mbit/s.

This still doesn't explain the test methodology. If you just used FTP to send file, it will not show true maximum. If your device reached 100% CPU, it means it was processing the traffic, so you must have some firewall or something.

Please tell us these things:

- what program was tranferring the data
- what data was transferred?
- how many connections
- TCP or UDP
- What configuation do you have on the router (paste output of "/export compact")

The real question is why does the device reaches 100% before it has maxed-out it's network speed? Seems a CPU bottleneck or something.

- I was using Windows file transfer mostly. This means that I had two computers connected to the router, each in it's own subnet so I could meassure transfer speeds over routed networks.
- Larger files were transferred 700MB+
- Only these two computers, no WAN or other network traffic, no firewall rules, routing protocols, etc.
- TCP seems to be the problem, tried also streaming and I did not reach 100%, but streaming does not take full advatage of the newtork line speed in my case
- I can paste it when I come home. But I really don't see the point as this router has been set to factory defaults before test and I only configured it with 2 IP addresses, each for a diffrent subnet

EDIT:

I also set a NFS server on 1 computer and created a NFS share on the other computer (changed the system from Windows 7 PRO 64-bit to ESXi 5.1u1 on 1 side and Linux Debian 7 x64 on the other side). Then I created a datastore on ESXi and linked it to the NFS share on the other computer. I created a virtual disk drive for a Windows 7 virtual machine on a newly created datastore and tried copying files into it. Again 100% CPU....

please upgrade to RouterOS v6.3 as there is a fix that will fix your problems:
*) RB2011L - fixed occasional gigabit switch-chip lockup;

It's very bizarre reading the above statement about the fix, while later on reading the following:

If the RouterOS v6.3 didn't help we suggest to try out this RouterBOOT firmware release:
http://www.mikrotik.com/download/share/ar9344_3_09.fwf

NOTE: after upgrading the RouterBOOT to v3.09 you can't downgrade the RouterOS to v5.x

To my understanding issue has not been yet fully resolved or at least not yet tested to be verified. Is this just a matter of time since v6.3 was released and routers don't have the appropriate uptime to conclude?

The fact is that I'm not so confident to spend money on buying a new device from these series even though I really like it.

It works for me. 5 days uptime since 6.4.

It works for me. 5 days uptime since 6.4.

have you confirmed again that version 6.4 is stable on RB2011LS. Because i am having a serious issue on this rb2011 and that is causing alot of problem for me. or do u suggest i use ether6 as my wan link and leave ether1 alone on this issue. Thanks

50 days uptime

same problem here

http://prntscr.com/3lrqd3

http://prntscr.com/3lrqfx

ports just go down only reboot helps but only for couple of hours ...

i found something but i just guessing

http://prntscr.com/3ls388

i have changed this option from "fallback" to "disable" because i different locations where is OK it was on "disable" we will see how it works

noooo it's not that again fail switch2 goes down

at least, I am not alone going mad. I have already replaced one with a brand new similar model thinking it was hardware but it isn't.

My router is now hanging up randomly so it can be hours or minutes before it goes down again.

Has any one been able to find a solution.

I inherited this setup from my predecessor so no choice but to fix it.

Hello,

Having this problem on my RB2011UiAS-2HnD, the router hangs, barely just can enter with winbox to restart, then works fine for some time.

CPU is about 10-30%, but when used rb750 with same config never hangs and no more than 8% cpu ever!


guys here talking about 6.4 packages, where can I get it? becaus eon offciial site its only 6.28 listed and many suggest that its still buggy on this hardware

any help appreciated

Why re-open this old topic? Your problems don't seem related except that you have an RB2011 which crashes.

Personally I run several RB2011's with switch, bridge, routing, NAT, etc. configured and have never seen this problem. I easily get uptimes of 150+ days. So that leaves me to think it might be configuration related? Maybe loops in the network?