A place to ask questions, connect with others, and stay in the know
I was looking at my calendar last week and I realized that we're already in Q4 of 2022. I can't imagine how quickly time went by – I guess most of you can relate. As the saying goes, time flies when you're having fun (or when you got a lot of stuff to do and deadlines to meet).
I remember that it's also around this time of the year when we usually notice a substantial uptick in interaction volumes, particularly when it comes to calls. I thought I'd just revisit a topic in relation to the influx of calls and that's scaling AWS Edges.
As most of you know, Edges are cloud-managed network appliances that manage and control the data flow between networks. It manages the connections to SIP phones, trunks, proxies, gateways, and IP PBXes. It also functions as a Phone Provisioning Server, IP PBX, Media Server, and Application Server for Genesys Cloud.
Customers will either have Virtual Edges for BYOC Cloud or Genesys Cloud Voice deployments or Physical Edges for BYOC Premise based telephony models (Sometimes both – specially for Orgs with Hybrid Media capabilities).
I'll focus on the former for this post.
Edge Scaling is the process of automatically, or manually, increasing the number of virtual Edges a Genesys Cloud customer has. Scaling usually happens when you need to increase your concurrent call capacity. This is the total number of calls you can have on your system at any one point in time. Basically, customers' Virtual Edges are hosted by Genesys in AWS. As mentioned above, this includes two deployment models:
It's important to note that edge scaling does not apply to Local Deployment Models (LDM) and Bring Your Own Carrier Premises (BYOC Prem) customers. These deployment types use physical Edge hardware, and such Orgs must purchase additional Edges to increase call capacity.
Genesys follows an Auto Scaling model. It detects when an Org needs more Edges. Once a need is detected, automated scaling then adds additional Edges to the Org without any manual input from a person. This 'auto-scaling' comes in two forms.
Based on concurrent call sessions on the Edge. When an environment needs to be scaled up, Genesys Cloud tooling adds another 50% of the current Edge count (so if the customer has 8 Edges, we add another 4 Edges), and balance across those. The environment is then scaled back down gracefully, after their spike is over.
ASG (Auto Scaling Group) scaling is temporary. What this means is that it does NOT change the "desiredCapacity" setting of the ASG so during the next auto-upgrade cycles any added Edges will be removed.
Allows an Edge to make a one-time call for a new Edge when its memory utilization hits 70% for 2min or longer. This scaling IS permanent as it adds +1 to the 'desiredCapacity'. This means that Edges added via Memory Scaling will be kept through auto-upgrade cycles.
If you feel that automated edge scaling won't be able to accommodate your surge with your current edge set up, submit a case with Product Support ahead of time to be able to accommodate the anticipated surge in your call volume.
Please ensure that the following are observed when submitting manual Edge Scaling requests:
I was looking at my calendar last week and I realized that we're already in Q4 of 2022. I can't imagine how quickly time went by – I guess most you can relate. As the saying goes, time flies when you're having fun (of when you got a lot of stuff to do and deadlines to meet).
I remember that it's also around this time of the year when we usually notice a substantial uptick in interaction volumes, particularly when it comes to calls. I thought I'd just revisit a topic in relation to the influx of calls and that's scaling Edges.
Please ensure that the following are observed when submitted Edge Scaling requests:
In most cases, new orgs are usually set up with two t2/t3 medium edges. A single edge has the ability to handle 250 concurrent sessions.If for example you have 10 edges currently deployed, you should be able to handle 2500 sessions. Its important to note that an active call will have 2 sessions or "call legs"- the first leg or session is from the carrier to the edge and the second leg or session is from the edge to the agent. The call will need these two legs to stand.Basically, with 10 edges you'll be able to handle 1250 concurrent calls.The automated scaling feature detects when a specific Org is in need of additional edges. Once a need is detected, automated scaling then adds additional Edges to the Org without any manual input from the customer and then after about 10-15 minutes, the system will run a query to see if further scaling is needed.What we want to avoid are instances where there is a massive surge of calls and the speed in which the calls increase outpaces the scaling capabilities of the edgeWhen you have a high call volume forecast and a case is opened with Product Support, our team will look at your forecasted concurrent call volume, and divide that number by 1/2 of an edge's capacity (forecasted volume / 125). The result we get out of that equation is then rounded to the next whole number which will be equal to the number of edges that will be needed.Example: Forecasted Concurrent Volume: 800 / 125 = 6.4 - we round this to 7 and that is the number of edges needed. As a best practice we add two additional edges to provide buffer for scaling and that will put you at 9 edges.
For the number of users you have, 2 edges seem low (I am only assuming that only medium sized edges were deployed when the org was set up). The edge will start to scale when you're down to the last 20% of your remaining edge session capacity and the Genesys Cloud system will check every 10 to 15 minutes to see if you'll need to scale further.
Lets say you actually hit your max edge capacity before the system is able to do the next scale cycle, no additional calls will be able to get in until active sessions are freed up.As you've stated, there is not a way for customers to look up the information regarding the edge type initially deployed on your org. The edge sizes Genesys Cloud uses are as follows:
During customer Onboarding for Genesys Cloud, there is a Care Gate document that is gathered from customers before initial go-live. There is a question (question number 25) regarding the max number of concurrent calls expected which is given to a team in product support that can check the information provided.
The intent is for all customers to have that doc vetted by product support in order to proactively adjust edge count/edge types (if necessary) before the customer's actual Go Live date.For customers that are already live in their production environent, the best way to determine the edge scaling behavioral patterns / trends for your organization is to open a case with Product Support so that they can provide a recommendation and effect the necessary adjustments on the number (or type) of Genesys Cloud edges based on their findings.
Hiya James!You're most welcome :)Thanks for sharing that piece of intel. I was actually thinking about that yesterday and looks like your edges should be able to take on some relatively moderate spikes in volume.
Every year, Genesys® orchestrates more than 70 billion remarkable customer experiences for organizations in more than 100 countries. Through the power of our cloud, digital and AI technologies, organizations can realize Experience as a Service℠, our vision for empathetic customer experiences at scale. With Genesys, organizations have the power to deliver proactive, predictive, and hyper personalized experiences to deepen their customer connection across every marketing, sales, and service moment on any channel, while also improving employee productivity and engagement. By transforming back-office technology to a modern revenue velocity engine Genesys enables true intimacy at scale to foster customer trust and loyalty.