Modern Colocation and Managed Infrastructure: Simplifying IT with XaaS

Edited for readability. Filler words, repeated starts, and incidental pauses have been removed. Timestamps refer to the original recording.

[00:00:02] Paul and Mike — Opening exchange:
Hello, everybody. We’re giving people another minute to join. We appreciate your time, and it’s good to see familiar faces and new ones.

[00:00:56] Paul:
We’re doing a series of these sessions, and we hope you find them valuable. The idea is to be more technical than sales-focused. We have a session coming up in September around HPE’s hypervisor alternative, and one in November around Nutanix and its hypervisor alternative. A lot of our customers are feeling pain with VMware, so we’re trying to give people alternatives.

[00:01:47] Paul:
A quick housekeeping note: I’ll be monitoring the chat. If you have questions, put them there, and I’ll interrupt Mike at natural breaks during the session. We’ll start with a welcome and total cost of ownership, then Mike will talk about 11:11. We’ll show some of the services we deliver with them and have Q&A at the end. I’ll hold questions until then where possible, but if something is important to address immediately, I’ll ask Mike to answer it at the time.

[00:02:33] Paul:
I like to think about things from a dollars-and-cents perspective and boil everything down to total cost of ownership. If you’ve worked with CPP and me for years, you know this. It’s always great to get the technology you want, but you still need to justify the cost.

[00:03:00] Paul:
This is a real-world example we’re working on. I’m happy to get into the details on a subsequent call or by text or chat. What we’re seeing is a 30–50% savings by going with infrastructure as a service. Infrastructure is expensive right now, including memory costs, and the cost of VMware is ballooning. I’ve heard some concerning things from VMware representatives over the last year, and I don’t anticipate their pricing stabilizing. Their goal is essentially to get 100% growth in every account every three years. Whether you grow or not, they’re looking to double your price.

[00:03:58] Paul:
This is a countermeasure to that. Customers looking to go to the cloud may still have on-premises infrastructure for things that need to run in VMs locally. This gives you an alternative. There’s a cost-savings story here. If you like what Mike has to say, the demonstrations, and the technical discussion, I can help you make the case within your organization. Thank you for joining. I’ll pass it over to Mike.

[00:04:35] Mike:
Thanks, Paul. VMware is the hot topic. How do we move to something scalable that gives us a good return on investment and a lower TCO without changing the way we do business? 11:11 is a great solution for that. We’ve been an 11:11 partner for quite a while—certainly longer than five years, probably six or seven—and we have a significant number of customers using the platform in various capacities.

[00:05:15] Mike:
I’ll start with some background on 11:11 and its features. Then we’ll look at the console and go through a prebuilt demo. That demo can be made available if you want to explore it on your own.

[00:05:46] Mike:
11:11 is a full cloud solutions provider backed by VMware, built on traditional storage and compute in a three-tier architecture. It’s built with data protection, disaster recovery, and business continuity in mind. Unlike moving to a hyperscaler such as Azure, AWS, or GCP, you can migrate or run your workload as it is currently configured in your infrastructure, in exactly the same way at 11:11.

[00:06:30] Mike:
You would be running on VMware with licensing included. Microsoft licensing for your workloads is also included, except for SQL licenses and Remote Desktop CALs. Backup and disaster recovery are integrated, and we can also integrate backup for Microsoft 365.

[00:07:00] Mike:
11:11 is built around a sound compliance infrastructure, as you would expect from a global cloud hosting company. That includes areas such as GDPR, CMMC, and ISO. If you’re an 11:11 customer through CPP, we can provide compliance documents. That gives you documentation to bring to your compliance team, insurance company, or other organization you report to.

[00:07:41] Mike:
On the security side, there’s infrastructure integrity monitoring, built-in antivirus and anti-malware, IDS, IPS, event logging, and two-factor authentication. A tremendous amount of security is built into the platform by default. This doesn’t prevent you from bringing your own firewall, adding security measures such as Arctic Wolf or SOAR options, or taking advantage of what we can do natively with 11:11.

[00:08:22] Mike:
The 11:11 infrastructure is built around NSX, including networking and edge firewall services. Every one of our current infrastructure-as-a-service customers brings their own firewall. That could be a virtual appliance, such as FortiGate, Palo Alto, or Cisco, or a physical appliance we colocate with the tenant. We have customers using those scenarios.

[00:09:05] Mike:
We can bring in circuits through 11:11 as well. If you need an EPL circuit or a dedicated site-to-site link, we can include that in the solution and the bill, so you don’t have to shop for carriers on your own. We can also help with carrier services independently. Whether you’re using infrastructure as a service, disaster recovery as a service, backup as a service, or object storage as a service, we can scope connectivity around the speeds and capacity you need.

[00:09:52] Mike:
From a support perspective, we have direct relationships with the 11:11 executive teams and meet with them regularly. We have dedicated customer success managers. Our NOC is tightly integrated with their NOC, with 24/7/365 support access. We handle tickets, case management, and escalations. You can sign up for real-time service and status alerts, and your portal provides reports and dashboards.

[00:10:34] Mike:
11:11 is built on familiar technology. It’s VMware-based infrastructure, with a roadmap to VCF 9. Veeam and Cohesity are built into the platform, along with Cisco switching and HPE storage and compute. Zerto is the data mover and business continuity platform we use to move between sites. If you’re a disaster-recovery-as-a-service customer, you’re most likely using Zerto to replicate into 11:11. If you’re running infrastructure as a service with 11:11, we use Zerto to replicate data between sites within its ecosystem. There’s also a strong relationship with Trend if you want to use the built-in security.

[00:11:29] Mike:
Let’s talk about the console, where we’ll spend most of our time. Everything you need for day-to-day infrastructure management is available there: operations, backup, disaster recovery, replication, building new servers, and networking.

[00:12:06] Mike:
Our NOC team can help manage this day to day. Under the shared responsibility model, the platform is 11:11’s responsibility. Its job is to keep the facilities available and the hypervisors running so we can run workloads on them. CPP can provide basic monitoring and alerting or full managed services, including day-to-day administration, change control, and best practices. We can also help with basic monitoring, alerting, and escalation.

[00:12:53] Mike:
Infrastructure as a service is designed around an all-in monthly cost. It’s as close to fixed pricing as you’re going to get with cloud consumption services. We try to size it for what you need, with a little headroom, so the bill doesn’t change each month. If you go over, resources are available and burst charges apply. They’re relatively affordable. If the usage is sustained, we’ll adjust the contract so you’re not continually bursting and fold those requirements into an upgrade.

[00:13:40] Mike:
Everything is built on SSD or NVMe. You have unlimited bandwidth, an unlimited number of VMs, a 100% availability SLA, and encryption. We’ve already discussed compliance. For protection, we can do cloud-to-cloud replication between sites in the 11:11 ecosystem and integrated Cohesity backups with immutability. Data protection and business continuity are built into the platform; we need to include them in the solution design or add them to an existing subscription. You can also use the security suite.

[00:14:28] Mike:
I want to emphasize the licensing component. You’re running VMware infrastructure with licensing included in the IaaS platform, and you’ve reduced separate Microsoft licensing to what you need for databases and RDS CALs. Those two differences add significant value. Looking three, four, or five years ahead, you would otherwise still be responsible for licensing renewals. This model addresses that. You also get the speed of cloud: if you need a workload or network, you can create it. If you need additional capacity, it’s a phone call.

[00:15:26] Mike:
We also offer the Flex Cloud platform. It provides the same type of infrastructure-as-a-service solution in a hosted private cloud, with dedicated compute. You can control vCenter and the hosts. It’s ideal for VDI and environments with significant compliance requirements, with the same level of support and more workload flexibility. VDI is a key workload we can’t run in the multitenant infrastructure-as-a-service environment, so this is where Flex Cloud comes into play. We can also support AI workloads, high-performance databases, and high-performance computing here.

[00:16:27] Paul and Mike — AI workload question:
Paul: That was one of the questions: AI workloads can be expensive to run in the cloud. What are the possibilities here? Mike: We can do that. There are standard sizes, and we can also do custom builds with GPUs. It becomes more like a colocation experience, but we can specify the hardware and build a more tailored solution.

[00:17:08] Paul and Mike — Management responsibilities:
Paul: Once it’s built, management of the infrastructure is on us and 11:11, correct? Mike: Correct. There are two ways to do that. We can get VMware licensing through 11:11 as part of the package, or you can bring your own VMware licensing, another hypervisor license, or run bare metal.

[00:17:41] Mike:
In that scenario, 11:11 is responsible for providing the hardware and ensuring the hardware platform is operational. The customer, or CPP under a managed solution, is responsible for patching the hypervisors and vCenter. We could also run HPE VM Essentials. Flex Cloud gives you options beyond Broadcom, including bringing another hypervisor.

[00:18:24] Mike:
We can move workloads between the infrastructure-as-a-service environment and Flex Cloud. We’re working with a disaster-recovery-as-a-service customer now, and one option we explored was putting its higher-performance workloads in a private Flex Cloud and allowing vMotion between the two. That’s possible if needed. Flex Cloud really is flexible.

[00:19:23] Mike:
We won’t go deeply into role-based access in the demo, but it’s important. If you have different tiers, roles, or functions on your team, we can use predefined roles or customize them. A manager who just wants to track usage could have a read-only role. Someone else could manage servers or access the console without changing the VM configuration. We can make those roles very granular.

[00:20:17] Mike:
At the edge, we can use NSX unless you bring your own firewall, which we generally recommend. You’ll likely want to match the policies, security profiles, and tunneling you already use on premises. As with public cloud, we can bring in a firewall and run it as a VM. We can also bring in a physical high-availability pair.

[00:21:09] Mike:
You can track general usage, see trends, and identify whether you’re bursting. This gives you a breakdown of what’s happening day to day. In the demo, we’ll build a workload, back it up and protect it with Cohesity, then build a protection group for the VM using Zerto.

[00:21:57] Mike:
We can build backups around the entity type or the VDC—the virtual data center. If a workload enters the virtual data center, it can automatically be backed up with everything else. If you need different schedules, we can get granular. We can create hourly, daily, weekly, monthly, or yearly backups, schedule snapshots, and configure extended retention. Policies can move backups to object storage for a nearline or offline scenario.

[00:22:48] Mike:
Analytics show how long backups take, whether they succeeded, and the data change rates. That helps with ongoing budgeting. If we’re bursting, we can investigate where the storage usage is coming from. We can restore files and folders or an entire VM, and download the files we need. You don’t lose that functionality. If you’re using Veeam, Cohesity, or Rubrik on premises, you get comparable functionality here. The advantage is that everything is in one console, so you don’t have to log in to multiple systems to manage your infrastructure.

[00:23:55] Mike:
For disaster recovery, this slide shows the site-to-site experience. This applies whether you’re replicating from an on-premises site or between 11:11 locations—for example, from Philadelphia to Scottsdale, Arizona, or Dallas. The setup is the same: Zerto and its Virtual Replication Appliances are deployed, then protection groups are built.

[00:24:37] Mike:
Inside the environment, you don’t need to manage those underlying components. You see a business continuity recovery group that you can build, test, and report on. You don’t have to manage Zerto day to day. It’s one of my favorite technologies and isn’t difficult to manage, but this is still one less thing on your plate. You can run a business continuity or disaster recovery scenario in your shared or private cloud with a few clicks.

[00:25:23] Mike:
Traffic is encrypted, and workloads are replicated in near real time using continuous data protection. Checkpoints average every five to ten seconds, so recovery point objectives are very tight. Through iterative testing, you can bring recovery time down to something reasonable, recover the organization within hours, and mitigate risks from ransomware, malware, and other threats.

[00:26:08] Mike:
We can build runbooks to recover in priority order or automate recovery with a click. The console lets you monitor recovery points, the last test, actual RPO, performance, change rates, priority, and the service profile. If you’re outside a five-minute RPO, you’ll start getting alerts. You also get detailed, step-by-step reporting. Customers with compliance reporting requirements can download a PDF report.

[00:27:22] Paul and Mike — Demo transition:
Mike: I’ll stop there for questions before we move into the demo. Paul: Please put questions in the chat along the way. We have dozens of clients successfully using this technology. If you’d like to speak with someone who has taken this approach and hear about their experience, let us know.

[00:28:00] Mike:
This is the console. We’re in a virtual data center in our lab, located in Reston, Virginia. You can see the data center name, associated organization, vApps—which are groups of virtual machines—and the VMs themselves. You can also see the edge gateway, essentially the edge firewall, plus private and public catalogs. The catalogs provide prebuilt images, ISOs, and VM templates. Recovery groups appear below.

[00:29:14] Mike:
We’ll start in the public catalog. There are already workloads available, including Ubuntu, Debian, CentOS, and Windows Server 2019 Datacenter. Here are two templates in the public catalog available to everyone: one with the Trend agent and one without. If you want the agent, it’s included in the price and comes as part of the operating system image.

[00:30:11] Mike:
The private catalog uses storage from our tenant. We’re using about 1.7 or 1.8 terabytes here. It lists our images, including a 7.0 ISO, a vCenter appliance, and a Debian image. If I want to upload a template, I can give it a name and description, then import an OVF or OVA for a virtual appliance.

[00:31:05] Mike:
We do this frequently with customers. One customer uses Security Onion to scan its environment. That customer works with our NOC team to upload the OVF and deploy it. The same approach works for virtual appliances or VMs you’ve exported and want to import without going through a migration.

[00:31:39] Mike:
I’ll use the Debian template here. I can deploy a vApp template, which creates a group of VMs and deploys a virtual machine into that group. I can also download the template or update its metadata. If I download it, that activity appears in the task center. Now we’ll go through a demo deployment.

[00:32:41] Mike:
We confirm the VM and its destination, then configure the hostname and basic settings: the name, NIC type, network connection, and IP address. Everything is stored on accelerated storage with advanced security, including encryption at rest. We review the configuration, confirm the destination data center, and click Submit.

[00:33:23] Mike:
The task list shows activity such as generating vulnerability or HIPAA reports, as well as the deployment we just submitted. We can see what the task was, who performed it, when it started, and when it completed.

[00:33:58] Mike:
Now we’ll go to our vApp and select the guest VM. We can reconfigure it. A VM from a catalog has some settings already configured, but we can change the name and description, number of vCPUs, memory, and storage. Guest customization options include changing the SID for Windows. We get a summary of the changes, then power on the VM.

[00:34:56] Mike:
This page shows whether Trend security is enabled and integrated with the 11:11 security functions, whether the VM belongs to a backup group, and whether it has a snapshot. Snapshots older than 24 hours trigger persistent alarms. You can see whether an ISO is mounted, along with VM telemetry: memory, storage, CPU, hardware version, tools version, running status, and IP address.

[00:35:39] Mike:
We can open the console. This is the actual VM console through vCenter, delivered through the same interface. You can control who gets console access. You can see how this process would work for a Windows VM, another Linux VM, or an imported OVF for a virtual appliance.

[00:36:25] Mike:
Next, let’s back up those VMs. Our dashboard shows that advanced backups are enabled, with three vApps and twelve VMs. We can see vCPU usage, assigned memory, and storage. The red numbers indicate usage above the reserve limits. Memory isn’t an issue, but we’re about 800 gigabytes over on storage and slightly over on CPU utilization. We can also see recent events and status information. We’ll go directly to advanced backups.

[00:37:17] Mike:
This shows the status of our backup process: how many jobs are running, how many times they’ve run, and how many runs succeeded. The information updates in real time. Below that are the policies and tiers. We’ll create a policy called “Advanced Backup Example.”

[00:37:56] Mike:
The available backup types include block-based, file-based, VM, vApp, VDC, RMAN, and SQL. I initially selected SQL Server; for this example, we’ll back up the VMs in the entire VDC. We can change the schedule using hours, minutes, days, or weeks. Data Lock provides immutability, and we can enable extended retention before submitting.

[00:38:51] Mike:
We can adjust the backup frequency, retention period, and lock period. For example, we can run a backup every eight hours and change how long it’s retained and locked. Extended retention is designed to offload backups to the 11:11 Cyber Vault, which is offline. This supports the three-two-one data protection approach described here: three copies of your data, two locations, and one offline copy. We won’t configure that option in this demo. Once submitted, “Advanced Backup Example” appears with a summary showing backups every eight hours, retained for four weeks.

[00:40:01] Mike:
Now that we have a policy, we’ll click Actions and create a backup group. We can back up the entire virtual data center, which includes every VM and workload in it. Alternatively, we can select individual VMs. If your environment includes lab, demo, and test workloads that you don’t want to back up, you can choose just the production VMs.

[00:41:11] Mike:
We assign a group name, choose a policy, set the time, and click Submit. We’ll use our example policy, but other policies are available. The backup group is now complete, and we can click “Run Now” to start the backup.

[00:41:56] Mike:
For workloads that require an agent, we can install one. For example, you might have a colocated server within 11:11 that must run on bare metal, whether through Flex Cloud or a colocated 1U or 2U server. We can push an agent to those servers. Options include Linux, Windows, Oracle databases, and Microsoft SQL. We provide a hostname or IP address, register the agent, and create a backup job.

[00:42:50] Mike:
Next, I want to show how we protect the environment. [The demo temporarily freezes.]

[00:43:54] Paul and Mike — Demo pause:
Paul: A couple of questions have come in. We could address those while the demo recovers. Mike: Before we get to those, I want to go back and show recovery.

[00:44:15] Mike:
Backup is good; recovery is better. I can recover files and folders by downloading what I need or restoring to the original location. We select a file and a recovery point from the available backups. The review shows what we’re recovering and the task name. I click “Create Bundle,” and once the task completes, I can download the files.

[00:45:15] Mike:
Alternatively, I can recover a virtual machine to its original location or a new one and select the vApp. I can disable automatic power-on so the VM doesn’t start immediately. I can add “recovery” as a prefix or suffix to the restored VM’s name, or use both. The wizard walks me through restoration before I submit it.

[00:46:16] Mike:
We recovered the VM to a vApp, and here is the recovery VM in our bubble vApp. I can power it on and make configuration changes where available. Once a VM is powered on, the changes you can make are limited, as in VMware generally. We can rename it, add memory where supported, or adjust the NIC. Finally, I can open the VM to confirm it’s accessible.

[00:47:17] Paul:
If someone decides to take the next step, how do you size what they need? What’s involved?

[00:47:33] Mike:
There are a couple of tools we can use. For VMware environments, 11:11 has a tool called Catalyst. It takes about five minutes to run. You download it, connect to vCenter, and select the VMs you want to run at 11:11. This applies to infrastructure as a service, backup as a service, and disaster recovery as a service.

[00:47:59] Mike:
Catalyst gathers telemetry already available in vCenter and consolidates it so 11:11 can complete the sizing and return it to us relatively quickly. It captures vCPU, memory, and storage utilization, along with VM geometry and configurations such as RDMs and iSCSI mounts. There are restrictions: RDMs aren’t permitted in the shared infrastructure-as-a-service environment, although we can accommodate them in private cloud. The tool also captures the guest operating system and unusual configurations. We can then price the solution. We can also use HPE CloudPhysics, which we use frequently, or RVTools to get started.

[00:49:01] Paul:
Thanks, Mike. Another question is about the contract. What commitments are required? Is it multiyear or single-year?

[00:49:17] Mike:
You’ll generally see a better discount with a multiyear commitment, but one-year contracts are available. There are minimum tenant sizes, but they’re low: a couple of terabytes of storage, some vCPU, and some memory. They’re fairly basic; someone with a file server could probably reach that minimum.

[00:49:48] Paul:
There’s a question about firewalls. Does it have to be a VM, or can you bring a physical device?

[00:49:55] Mike:
You can bring a physical device. We have customers colocating a high-availability pair of their own firewalls. Servers and backup appliances can also be colocated. One thing to understand is that 11:11 doesn’t make its vCenter server available in the shared environment. That means you couldn’t restore directly into that environment from a backup appliance you brought in through colocation.

[00:50:31] Mike:
Network devices are no problem. We need to know the port count, rack space, and power requirements. Ten-gigabit ports and power are included.

[00:50:49] Paul:
It’s worth revisiting the backup device point. We have customers using Rubrik or Cohesity who want a similar device in the colocation facility. Could that be a backup solution while using 11:11 shared infrastructure?

[00:51:14] Mike:
If they want to use that backup solution directly with their workloads, we should focus on Flex Cloud, the private offering. There, we can stand up vCenter or another hypervisor manager and connect it directly to a colocated appliance. In the shared environment, 11:11 doesn’t make vCenter available. A colocated appliance can still be useful for getting backups offsite from your facility. If you need to run workloads on the compute included with a Rubrik or Cohesity appliance, we can configure trunk links so you can access those workloads.

[00:52:12] Mike:
This is the dashboard with Zerto enabled. It shows how many recovery groups, runbooks, and replicated VMs we have, assuming those have already been configured. We can open the failover wizard to move workloads from one site to another and choose either a test or a live failover.

[00:52:50] Mike:
A test uses our failover test networks and places workloads in an isolated bubble. A live failover follows the configured requirements, including whether to retain IP addresses, change names, or run pre- or post-failover scripts. It can follow a runbook or recovery group. We’ll use a recovery group, choose the application, and adjust the checkpoint. I’ll select the 9:53 checkpoint.

[00:53:40] Mike:
The wizard warns that this is a live failover, so the workload will run elsewhere. We need to choose whether to commit it or roll it back. If we’re bringing up the workload as production in the disaster recovery environment, we need to shut it down in the original production environment. The networks are connected for production use, so leaving the original workload running could create a conflict. A test instead uses the failover test networks.

[00:54:20] Mike:
After submitting, we can watch the task progress in real time. Once the protection groups are defined, Zerto handles failover. At this point, it’s waiting for user input. We can open the CRM application after Zerto has failed over the VMs, brought them up, and checked that they’re running. The interface shows protection and performance information, and we can validate that the application is available.

[00:55:25] Mike:
The remaining information shows the test networks, target networks, public IP for replication, and where workloads will land if they fail over. We confirm that the result worked, then click Commit. Once it commits, we can download the report.

[00:56:18] Mike:
The detailed report identifies the protected site, recovery site, where the workload landed, any guest customization, and the recovery settings. In your environment, there would be more information populated. It documents the entire process, from shutting down the protected VM to bringing it up and confirming it worked.

[00:56:47] Paul and Mike — Closing:
Paul: Thank you, Mike. We’re approaching the top of the hour. Thank you, everybody, for joining. If you’re interested in learning more, let us know. We’re happy to get the Catalyst tool running in your environment. Have a nice day. Mike: Thanks, everyone.