Exam AZ-300 topic 2 question 42 discussion

BOX 1: NO BOX 2: YES BOX 3: YES

Box 2: No - Web requests uses the TCP protocol, not the UDP protocol. The question asks about Rule1 and not Probe1. Box 3: Yes - VM2 will be deemed unhealthy which means all traffic will be routed to VM1

correct answer is NO YES YES mentioned on https://vceguide.com/when-you-request-a-webpage-named-page1-htm-from-the-internet-lb1-balances-the-web-requests-to-vm1-and-vm2-and-you-receive-a-response/

Given answer is correct. For Box3, even if html is deleted, still need 40seconds to 60 seconds for LB to deem VM2 unhealthy. Within this period, the web requests are still routed to VM2.

There is not UDP in the image. LB can probe either TCP/HTTP/HTTPS. TCP probes port, http/https probes pages. If the config is http and page was delete. The probe will get a 404 and wont route to that VM. https://docs.microsoft.com/en-us/azure/load-balancer/manage#health-probes

The correct Answer in the BOX3 is NO.. the statement says "LB1 will route all the web requests to VM1" and only the new requests will be sent to VM1 not all the requests.

Box 3: Depends on how you interpret the question. "LB1 will route all the Web requests to VM1". Are we talking about new Web requests or existing Web requests? YES: "When a health probe fails, Load Balancer will stop sending new flows to the respective unhealthy instance. Outbound connectivity is not impacted, only inbound connectivity is impacted." NO: "If a backend endpoint's health probe fails, established TCP connections to this backend endpoint continue."

No, Yes, Yes

Box 1 No - for obvious reasons. Box 2 No i saw the below info: Application payloads are transparent to the load balancer. Any UDP or TCP application can be supported. https://docs.microsoft.com/en-us/azure/load-balancer/concepts Box 3 No (https://docs.microsoft.com/en-us/azure/load-balancer/components, search for “threshold”) You can define the unhealthy threshold for your health probes. When a probe fails to respond, Load Balancer stops sending new connections to the unhealthy instances. A probe failure doesn't affect existing connections. The connection continues until the application: • Ends the flow • Idle timeout occurs • The VM shuts down Feel free to correct me.

Box 2 is definitely no. For those that think you can get HTTP working over UDP/80 I challenge you to do so and report back your progress ;). Box 3 is YES and I have confirmed in a lab. The load balancer is looking for the file /probe1.html on both vm1 and vm2, or more specifically an HTTP/200 response when it checks for the probe file. If you delete it from VM2, then it will return a 404 and mark it as down which means all web requests will go to VM1.

Box 3: No - Probe down behaviour - TCP connections New TCP connections will succeed to remaining healthy backend endpoint. If a backend endpoint's health probe fails, established TCP connections to this backend endpoint continue. https://docs.microsoft.com/en-us/azure/load-balancer/load-balancer-custom-probe-overview#probedown

Box 1: No No session persistence set up Box 2: Yes UDP isnt suported protocol for HTTP on a load balancer Box 3: Yes(but not positive) All I could find is a google cloud explanation, I would imagine its similiar if not the same for azure cloud--Specify the URL path to which Google Cloud sends health check probe requests. If omitted, Google Cloud sends probe requests to the root path, /. The request-path option doesn't support query parameters

this is regarding BOX 3 , i Have tried the scenario with AZ scale set , deleted the default.html and the traffic still going though and instead of the default page I have received the IIS page that implies that NO for Box 3 is correct 100%

The Health probe here is used by LB Rule1 which pings both VM1 and VM2 for health probe. However here we are deleting the whole health probe and rule2 will not find that health probe and will send 403 or some other response and will consider that both VMs are unhealthy. I think it will route traffic to VM1 if the health probe will be separate for both VMs and if one probe fails it will send traffic to another. It's a tricky but answer lies in carefully looking at setting. A Lab test will validate the correct assumption.

No, No, Yes

NO NO YES

N Y Y - Is port 80 TCP or UDP? In any case, though, TCP and UDP ports are separate. You can run an HTTP server on TCP port 80, while running something else on UDP port 80. One reason why someone might choose to do so is that TCP and UDP port 80 are both officially assigned to HTTP, even though HTTP-over-UDP is almost never used.