I’m not totally sure, but I’d pick B - the queue should be limiting the download speed for 192.168.1.10 to 1 Mbps. Simple queues apply limits based on direction and target, and it’s easy to confuse upload vs download here. If the image shows dst as the target, that usually means download. Anyone else think otherwise?

Its B for sure, exporting just the /ip firewall filter section is the cleanest. But are we allowed to use scripts or only the built-in export command here? If scripts aren’t allowed, it might limit some approaches.

Pretty sure it's D and E. Ping uses ICMP and traceroute commonly uses UDP packets to probe the path, so those fit. I don't see how DHCP or TCP would be correct here.

Yeah, it's definitely B. When 'check-gateway' is on, RouterOS keeps the unreachable gateway out of round robin and only uses the working ones. Been like that in every lab I've run. Pretty sure this is what they look for on the test, but open to corrections if anyone's seen a weird exception.

Yeah, this matches what I've seen in practice. B is correct, since with 'check-gateway' enabled, RouterOS automatically avoids sending packets to gateways it can't reach. Saw a similar question come up in some exam reports recently.

Is IEEE even a recognized ACL type anywhere or just trying to trip us up?

Feels like C is right, since 20 bits for hosts should be 2048 addresses. Maybe I'm missing something subtle about how "usable" is counted here, but that's what I'd pick based on the math. Open to a different take though!