Exam SY0-501 topic 1 question 496 discussion

I work with load balancers and have passed specialist exams in the subject. I don’t think there is enough information here to come to any sort of conclusion for what algorithm is best. Maybe some of the scenario is missing. But based on why I’m reading, answer is D. A - round robin - simply distributes connections over each server one at a time in order. No performance gain. B - weighted - will distribute connections depending on what services are running on the box and resources available. For example, if we have a pool of two servers which load balance for app A: - server 1 is running app A and SQL server. - server 2 is only running app A and happens to have a faster processor. On each connection distribution cycle we would then distribute 2 connections to server 2 and 1 connection to server 1 (its working harder than server 2 as it runs sql server and has less memory). Minimal performance gain. C -least connections - the servers with the lowest amount of connections will get the next incoming connection until it roughly balances out. No performance gain. D - based on geography, the closest and fastest responding server will receive the next connection. Good performance gain.

in my opinion that redundant servers must means those servers have duplicated configurations and they are used as backup servers in other hand they are logical servers. Which comes to conclusion that they are logical server/ load balancers.

It can't be D by any means. Locality based is well explained here https://www.istiobyexample.dev/locality-load-balancing In the question we don't have any geo locations mentioned. In the other hand please take a look here: https://docs.citrix.com/en-us/citrix-adc/current-release/load-balancing/load-balancing-customizing-algorithms/leastconnection-method.html Answer based on question asked is C

I guess this question is for Net+ not Sec+ , I see no relation with security here but performance.

The answer might be 'D' (Locality-based), but that's outside the scope of the exam. If the question is on the exam, it likely won't be graded so, f**k it -- I'm going with 'C' (Least Connections).

I believe the provided answer, D. Locality-based, is true. After some research i realized that, among existing load balancing techniques, Least Connections is the optimum for the servers which reside in the main site that the question implies. (https://www.liquidweb.com/kb/load-balancing-techniques-optimizations/). However, since "The applications are hosted on multiple servers and must be redundant." it is better to have another site for the servers to be geographically redundant. This is the point where locality-based technique steps in. The load balancer that the technician is configuring has "Locality-Based Least Connection Scheduling" algorithm. (https://blog.selectel.com/load-balancing-basic-algorithms-and-methods/). With this technique it is possible to perform optimum load balancing with merging both Least Connection and Locality-Based techniques (https://istiobyexample.dev/locality-load-balancing/).

B Weighted – Question says “to accelerate the network PERFORMANCE “ - [“weighted” can apply to both Round-Robin and Least-Connected and improves PERFORMANCE.] There are OTHER VERSIONS of the Least Connections algorithm, namely LOCALITY-BASED Least Connection Scheduling and LOCALITY-BASED Least Connection Scheduling with Replication Scheduling. Locality based load balancing only works when the caller has a Service associated with it. This is because it's locality is set based on the locality of the first Service. You shouldn't need a Service to do this -- think about situations where you are just a client, not a server as well. Only Istio defines locality based load balancing based on geographic location. Most other providers do not. Istio is an open source service mesh platform that provides a way to control how microservices share data with one another.

I'm going with C, never read about locality-based anywhere and it isn't in the objectives.

locality based isnt an objective i dont understand this

C Least Connected – “to accelerate the network performance “ - There can be instances when, even if two servers in a cluster have exactly the same specs (see first example/figure), one server can still get overloaded considerably faster than the other. This can cause the total current connections in Server 2 to pile up, while those of Server 1 (with clients connecting and disconnecting over shorter times) would virtually remain the same. The Least Connections algorithm takes into consideration the number of current connections each server has. When a client attempts to connect, the load balancer will try to determine which server has the least number of connections and then assign the new connection to that server. [NOTE: “weighted” can apply to both Round-Robin and Least Connected and improves performance.] https://www.jscape.com/blog/load-balancing-algorithms

It's C (Least Connection). The key indicator is 'accelerate the network performance'.

Geography isn't mentioned in the question. So to maximise performance, the application team would want to establish each connection with the server running the least number of connections. Hence I chose C.

It is locality-based because of the load-balancers are balancing based on location. Since the client has less distance/hops the client will get better performance.

this is out of the text-book and test range, hell but, yes, it should be D: Locality-prioritized load balancing Locality-prioritized load balancing is the default behavior for locality load balancing. In this mode, Istio tells Envoy to prioritize traffic to the workload instances most closely matching the locality of the Envoy sending the request. When all instances are healthy, the requests remains within the same locality. When instances become unhealthy, traffic spills over to instances in the next prioritized locality. This behavior continues until all localities are receiving traffic. You can find the exact percentages in the Envoy documentation.

If by locality based, you mean source affinity then why not call it source affinity? Some load balancers use source address affinity to direct the requests. Source affinity sends requests to the same server based on the requestor’s IP address.

I do not know why it is D. Locality based, should it not be A round robin ? There is no indication in the question of any reason to use anything other than normal round robin.