################### 
MY ENGLISH IS BAD, SORRY ########################
Hello guys. 
Thanks in advance for wasting precious time reading my question. I have a RB450, RouterOS v6.31, set to be a small FW company that has an average of 100 users simultaneously surfing the Internet.
Unfortunately it is restarting, as a single condition and for that I need to explain my scenario:
I use three interfaces of the equipment.
The interface 1 is my first ISP, call the ISP1, link ADSLv2 + with 90 Mb of Down and 30 PU. Dial-up connection PPPoe.
The interface 2 is my second ISP, ISP2 with 100 MB of Internet Down and 10 PU.
In my interface 5 is my local network, I call LAN.
Before you begin, I need to mention that this is my second conf what I do and that presents the same problem here in my structure. The first setting I left was pretty full and heavier than it is now. There was QOS rules, Mangle and scripts. But when we add this new ADSL link, had trouble climbing a PPPoE interface and only worked when I enabled TCP MSS Change on the ppp profiles. It would be suspicious because of PPPoe, I put a router in front of FW to turn this pppoe connection provider and released in the DMZ all connections to IP FW.
Unfortunately nothing. Keeps restarting.
Then went to my second configuration. I made it simpler, do not mangle, QOS anything. So much so that ISPs put my interfaces in brigder and Netwach do Failover. But the same thing happens when I turn on the ISP1 after an hour or less my RB restarts again. Unfortunately I can not a log that helps me in finding this problem. Luckily I managed to capture this error:
Interface, warning ether5-master: tx on loop interface (1)
Interface info Ether1-isp1 link down
Interface info Ether1 isp1 link-up (speed 100M, full duplex)
Note: Unfortunately my English is very bad, Iām not nation with the language and the fact that living in another country. Thank you for your attention and sorry guys.