Hello, I’m using an RB1200 very happy for the last year as my border router, I’m running several mangle rules and a large and very beatifull queue tree to priorize some services.
All works great, queue tree and mangle rules based on QoS and protocols works great.
Due to traffic growing, we decided to switch to a more powerfull router (~80% CPU usage in my old RB1200), here decided to migrate our border router to a new dual ppc RB1100AHx2 borad.
After migrating problems begins, the first I see is a heavy traffic drop, and then I see about 20~30% packet loss when I try to go trought this new router.
Severals try & error tests show me that if I disable the whole queue tree Packet loss drop to 0% and traffic grow up to the normal levels, ( of course I need queue tree to priorize traffic. ).
I see some issues with differents queue type when the platform is SMP, I tryed with all queue types and with all I have problems.
I see peoples having different problems with this router (RB1100AHx2) but not exactly what I have.
I’m using two ports of the router, ETH1 is my upstream and ETH6 is the downstream interface.
My RB1100AHx2 with RouterOS 5.21 began rebooting at random intervals when I pushed traffic through port Ether2. I don’t know if the port was important or just the switch group or if it was unrelated… MikroTik support gave me 6.0rc6. That ran good with traffic on port Ether2. Then I enabled my QoS queue tree again. Reboots. So MikroTik support gave me 6.0rc7. That is working with queue tree.
So, what I’m trying to say is, you may need 6.0rc7 or newer to make the RB1100AHx2 behave.
I wish not. I don’t like running production hardware on alpha level software. But when the alpha level software works better than released software, you do what you have to do. I am encouraged by the rapidity with which MikroTik support responded to my problem reports.
Unhappy with the change, We decided to roll back to the RB1200 Router.
RB1100AHx2 don’t work, packet loss when queue tree enabled.
We’re very unhappy with this hardware and we can’t find any response from mkt (we have tested diferents RB1100AHx2).
We see the same thing. There is clearly a problem after 5.6 (v6 also has the issue). You can install 5.6 to clear up the packet loss in my testing. I submitted a bug report with exacting reproduction steps back on dec 18th. Got initial run-around but maybe they are starting to look more closely. You should build your own support ticket to get it more attention.
5.22 here-- Put a new 1100AHx2 into production yesterday, discovered the hard way this morning that my queue trees were utterly obliterating VoIP audio quality. Good(?) to see I’m not the only one.
Random packetloss up to 10%, not working queue tree.
I had those issues with queue tree (300+ rules per interface, 2 interfaces) attached directly to ethernet interface (this setup in fact replacing interface-queue). Then i reworked qos to use global queue and the problems disapeared
I’ve tried the two ideas on changing the queue type and they did Not help. Using Simple queues greatly restricts the functionality of the device and is not ‘really’ a solution.
This problem can be reproduced on the bench 100% of the time IMHO. I’ve opened a ticket with support but have not heard back on it yet.
This is a pretty big issue for the 1100AHx2. Any other ideas greatly appreciated.
