VMware vCloud Availability for Cloud-to-Cloud DR – Pairing and Usability (2 of 2)

Continuation of VMware vCloud Availability for Cloud-to-Cloud DR – Pairing and Usability (1 of 2)

This post covers pairing and usability of VMware vCloud Availability for Cloud-to-Cloud DR.

Previous blog posts:

Blog Post – what is vCloud Availability for Cloud-to-Cloud DR? 

Blog Post – vCloud Availability for Cloud-to-Cloud DR – Installation and Setup

  1. Part 2:
    1. Initial authentication between the two sites
    2. Migrating a workload in the same site
    3. Protecting a workload between two sites
    4. Testing the protected workload
    5. Edit Options
    6. Failing Over

Initial authentication between the two sites

  1. This is pretty simple and very similar (if not exactly) like vCAv-DR2C. Go to Paired Clouds -> click the actions button to authenticate on the paired remote site. For me, this is SiteB. 
  2. We need to provide the organization name, the username of the org admin, and password. Note this must be an organization admin for pairing. 
  3. Complete! 
  4. Now, I can do the same on my second site, SiteB. 

Migrating a workload in the same site

  1. So let’s work through a use case where I want to migrate a vApp or VM between two different oVDCs in the SAME vCloud Director instance. This question came up on the vExpert Slack channel and I thought this was supported, but wanted to make 100% sure. The answer is yes – fully supported. So let’s go through how I made this happen.
  2. From the DR Workloads tab, let’s click on the Discovery button – 
  3. From here, our source is going to be our same site, which is SiteA/Org1, which should show host in parenthesis. 
  4. Let’s select my test vApp, which is properly named “vApp_test” for this exercise. 
  5. Now our destination is the same vCD instance, or SiteA/Org1. Let’s select it –
  6. We can now see the other org VDC available, which is my Org1-Gold-oVDC. This is the only selection available as we cannot move it to the same oVDC and I do not have more than two oVDC’s inside of this organization. From here, we can also select the Storage Profile, Target PRO, and if we want to add in any Point-in-Time instances along with data connection type. 
  7. Once we click finish, we get our setup screen… 
  8. We can now see the vApp is being configured and in the “Protecting” status. 
  9. Complete! All green and in the “Protected” state so it’s ready for migration over to my destination oVDC. 
  10. Let’s go ahead and click Failover and get the confirmation window on what we want to do. I can select the DR Network (I didn’t set up networking for this test) and if I want to turn on the target VM. 
  11. We can now see the status has changed to “Failing Over” – 
  12. From my vCD instance, I can see the vApp being imported… 
  13. Source/Original vApp is being powered off.. 
  14. Under Tasks on the vCAv portal, I can see the migration tasks underway – 
  15. Complete! We have successfully migrated over the vApp to the new orgVDC. We also see the original/source vApp was completely powered off, or in the Stopped state. 
  16. In conclusion, very simple and intuitive to migrate between the same vCD instance. Theoretically, you could deploy this at a single site and use this for local migration.

Protecting a workload between two sites

  1. So let’s cover migration between the two sites – SiteA and SiteB. In this exercise, we will be protecting a workload in SiteB and protecting it to SiteA. This is very similar to our exercise above (migrating between the same site) while we are selecting the paired site for protection.
  2. Let’s go ahead and click on Discovery and select our source site (SiteB) – 
  3. Now we can select our vApp that we will be protecting to SiteA – 
  4. Select our destination, which mine is SiteA/Org1 – 
  5. From our final screen, we can select the appropriate oVDC, storage profile, and my target RPO. For this exercise, I’ll be adding in some Point-in-Time instances too. Click OK and let’s get to protecting…
  6. I tried to grab the transition log, but it was too fast. But I do see the initial replication succeeded along with my protected vApp showing “Protected” – awesome! 
  7. I also like the event pop up we get when something changes. We can see this in the above screenshot that shows my newest protected vApp is good to go.

Testing the protected workload

  1. Testing is the ability to bring up the protected workload at the destination site in an isolated network of your choosing – this allows the application owner to verify everything is operational and could be used for regulatory purposes too.
  2. Testing is pretty easy – it can be orchestrated from either site (source or destination) and it’s with a click of a button –
  3. We get the confirmation screen and the choice of our test network we want to utilize. Once I hit the Start button, I can see the status changes to “Failover Testing Initializing.” 
  4. On my SiteA, I can see within the logs the failover testing is underway while we have a transition of a vApp inside of SiteA – 
  5. Alright, now it shows Failover Test Ready which is great. Now, my app owner can test their app on the destination and verify functionality. 
  6. One of the nice additions is the quick launch buttons on the test workload – we can hover over the two icons and see quick launch buttons to get to each site. Very nice addition. 
  7. Finally, when our testing is done, we can click the Cleanup button to remove the test VM and go back to our normal, Protected state. Pretty straightforward. 

Edit Options on protected workload

  1. Clicking the Edit button provides us with the ability to make changes to a current protected workload –
  2. From the options pane, we can see the following – 
  3. We get our standard RPO slider – from 5 minutes to 24 hour RPO selection – while providing the ability to keep point-in-time instances for further retention.
  4. Moreover, all traffic is encrypted but we can also further optimize by compressing data. Very similar to vSphere Replication, the replicators will attempt to compress the data to minimize network traffic.
  5. Last of all, we have the ability to quiesce the operating system by using VMware Tools.

Failing Over

  1. In our last example, we will fail over my core-B vApp from SiteB to SiteA. Failover can be done from either site (especially important if I lost my source site) and very straightforward.
  2. Let’s select Failover from the UI – 
  3. As discussed before, our standard options and what network we want to select. 
  4. We can see it transitioned to Failing Over… 
  5. Voila! Failed over. Now we can click the quick launch shortcut and start doing whatever we need to do. 
  6. Another thing to note – even with a failed over workload, we can reverse the replication and reprotect it back to our original site, assuming the site is still operational. This is done by selecting the Reverse button. Now, this will show as outgoing from my SiteA to SiteB.

That’s it, folks! My hope is this was informational for any providers that are considering to utilize Cloud to Cloud for migration and DR needs for their multi-site vCloud Director environments. It’s a great tool and is very intuitive for our tenants and providers.

Thanks!

-Daniel

VMware vCloud Availability for Cloud-to-Cloud DR – Pairing and Usability (1 of 2)

As a continuation of my vCloud Availability for Cloud-to-Cloud DR (vCAv-C2C) series, we will be covering the initial pairing and usage of C2C in this blog post.

Blog Post – what is vCloud Availability for Cloud-to-Cloud DR? 

Blog Post – vCloud Availability for Cloud-to-Cloud DR – Installation and Setup

I will break this post into the following sections and into two blog posts:

  1. Part 1:
    1. The pairing of the two vCD/vCAv sites
    2. Logging into the vCAv Portal and Portal Overview
    3. Provider view of the vCAv Portal
  2. Part 2:
    1. Initial authentication between the two sites
    2. Migrating a workload in the same site
    3. Protecting a workload between two sites
    4. Testing the protected workload
    5. Edit Options
    6. Failing Over

The pairing of the two vCD/vCAv Sites

  1. First, the Provider must set up an association between the two vCAv sites before the tenant can authenticate within the vCAv portal.
  2. This is done in the Replication Manager portal – point your browser to https://<vcav-rep-mgr>:8046, authenticate, and click on Sites -> New site
  3. We then assign a Site name, the full URL of the respective peer replication manager, and the appliance password. Within my setup, I did this on both sites, SiteA and SiteB. 
  4. Once we hit OK, you will receive a Task succeeded message – 
  5. Now, if we click on the Sites subsection, we can now see our local site along with our newly peered site. From SiteA and SiteB, I can see each respective site. 
  6. Complete! Now we are ready to log into the vCAv Portal.

Logging into the vCAv Portal and Portal Overview

  1. Point your browser to the fully qualified domain name along with port 8443 – this is the default port the portal runs (can be changed). 
  2. Logging in is very similar to vCAv-DR2C – you utilize the username@org-name parameter. For my SiteA, I have an “Org1” while my SiteB is “Org2” – for SiteA, I am logging in as org1admin@org1 with the appropriate password.
  3. Once authenticated, I get the very clean interface that shows the Cloud Topology. We haven’t started any replications/migrations yet, so I don’t have any ingress/egress traffic yet.  
  4. On the home page, you get a very clear view of the topology, workload statistics, and even the Organization VDC status. I thought this was a very efficient use of current oVDC utilization along with what’s being used within each oVDC. 
  5. Under the DR Workloads tab, we would be able to see incoming and outgoing protected workloads. Since we haven’t set up any yet, nothing to show yet. 
  6. Paired Clouds shows available vCD/vCAv instances and their authentication status. Since the Provider paired with SiteB, we have it available but it shows unauthenticated. We still need to authenticate with the appropriate credentials that reside on SiteB. 
  7. Last of all, Tasks will show any previous or current tasks and the event log. This will show logs for any connected site, so very easy to see exactly is going on between the paired sites. 
  8. For those of you that have used vCAv-DR2C, this is very similar to that experience with a few minor enhancements. The learning curve is very minimal and we will go through a few test scenarios.

Provider view of the vCAv Portal

  1. The development team did a great job from a provider view on providing very useful information.
  2. First of all, we get to see a current status ticker that shows state of the vCAv environment. I thought this was extremely useful and intuitive to gain an operational understanding of the health of the current environment. 
  3. I didn’t mention this in the previous portal post, but this is also available there too. You can expand and see the details of the current replications/migrations between sites. This is done by clicking the carrot icon in the top right corner. 
  4. If you scroll down also, you get to see the pVDC view of the current environment with a resource utilization rollup. Very similar experience to what we see now in vCD 9.1 Multi-Site. 
  5. Under the Orgs tab, we get to see the current organization and any registered vApps and current status. 
  6. One very nice item is that shield/checkmark icon. If we click that, we can actually impersonate the organization admin and assist with any tasks that the organization may be having trouble with. All done through this single interface! 
  7. I won’t cover DR Workloads as it’s the same as the tenant view, but just a macro-rollup of all protected workloads. However, under Administration, we get to see Configuration and Registration. From here, we can actually register other components to monitor from this dashboard along with resource threshold display. Very nice! 

Onto the next post, where we will review migrations and protection operations. Thanks!

-Daniel

VMware vCloud Availability for Cloud-to-Cloud DR – Installation and Setup

As discussed in my introduction post about vCloud Availability for Cloud-to-Cloud DR (vCAv-C2C), I am going to do a high-level write-up of my installation/deployment process for this new vCD site-to-site replication and migration solution.

While this write-up does cover the deployment methodology, a production state will need to be rationalized inclusive of certificates, ports, and firewall rules that will be required.

Below are the high-level steps I took for deployment. VMware’s documentation does go in a different order, but this is the way I rationalized it along with a combined role (all-in-one appliance).

High-Level Steps for Installation and Setup:

  1. Download OVF and deploy at each respective site
  2. Configure vCAv-C2C Replication Manager
  3. Configure vCAv-C2C Replicator
  4. Pair Replication Manager with Replicator
  5. Configure vCAv-C2C Replication Service/Manager
  6. Configure vCAv-C2C Availability Portal

Again, my high-level architecture –

vCenter Deployment

  1. Let’s grab the OVF (remember, we only need one for vCAv-C2C), and deploy it to my management cluster – 
  2.  We can see it’s the new Cloud to Cloud appliance – 
  3. As we discussed before, the team has done a great job of simplifying the deployment of vCAv-C2C to a simple and single OVF for all of the roles required for Cloud to Cloud. We can see the drop-down for each respective role. For my deployment, we will be selecting a combined since this is a lab environment. 
  4. We now have to put in the required network configuration while enabling SSH (this is not a mandated requirement, but nice to have in my lab environment). One thing to note – the root password is temporary. When we get into the initial portal configuration, it will prompt us to change it to a new password from a security perspective. 
  5. Final screen to complete the deployment – 
  6. Alright, initial appliance deployment is done! Now off to the configuration.

Configure vCAv-C2C Replication Manager

  1. For the initial configuration, we will want to open a browser to “https://Appliance-IP-address:8044” or in my case, https://vcav-repmgr-01a.corp.local:8044 – 
  2. Let’s click on the Configuration Portal link and now set the new password after a successful login with the password we set in the OVF deployment – 
  3. We can see we have a red x where the lookup service is missing. This is for us to point to the resource vCenter (remember, resource and management vCenters must be in the same SSO domain at each site). 
  4. Let’s go to Configuration -> Set lookup service and put in the lookup service FQDN and we should see a successful message. 
  5. Back to the diagnostics tab, we can now see the lookup service is green!
  6. Now, the next step is correctly set up the replication instance and then come back to the Replication Manager for pairing.

Configure vCAv-C2C Replicator

  1. Open a browser to “https://Appliance-IP-address:8043” or in my case, https://vcav-repmgr-01a.corp.local:8043 – 
  2. Change password – 
  3. Set lookup service – again, I am using a consolidated vCenter instance for resource and management so this is pretty straightforward – 
  4. Success! 
  5. Now, back to the Replication Manager for pairing.

Pair Replication Manager with Replicator

  1. On the Replication Manager, let’s go to Replicators -> New replicator while giving the full FQDN along with the password (remember, new password!) and SSO credentials –
  2. Accept the cert and we should see the success message shortly.
  3. Now, let’s click on Show all managers and we can see the replication instance is now registered – 
  4. While from the replication instance, we can see the replication manager also configured – 
  5. Excellent! Now, off to the Replication Service Manager

Configure vCAv-C2C Replication Service/Manager

  1. Open a browser to “https://Appliance-IP-address:8046” or in my case, https://vcav-repmgr-01a.corp.local:8046 – this time, I used the new password as it seems to have propagated to the other roles. 
  2. Now, we get a nice clean wizard for the setup. 
  3. Let’s put in the Site information, this is going to be my SiteA. 
  4. Next, lookup service again with accepting the certificate – 
  5. Now, we are going to point it to the Replication Manager we previously setup. This is on port 8044, so let’s put that full FQDN along with port 8044 – 
  6. Now, we get to setup vCloud Director. I used the manual configuration here to verify everything was good to go. 
  7. Finally, we get a summary screen to show the stated configuration – 
  8. Excellent! All green and expanding out the Manager data shows our registered replication instance too. Very slick interface. 
  9. Now, off to the Availability Portal configuration as the last setup step.

Configure vCAv-C2C Availability Portal

  1. Home stretch for the first site installation and configuration – let’s open a browser to “https://Appliance-IP-address:5480” or in my case, https://vcav-repmgr-01a.corp.local:5480 
  2. I am prompted for vApp Replication Manager / vCD Connection information. By default, I did see a “127.0.0.1:8046” but decided to change it to the FQDN of my combined instance. Put in the vCD credentials (administrator@system) and hit connect to get a successful message – 
  3. Now, click Test to verify everything is operational from a vCD perspective – 
  4. We will be using the defaults for the database, but you are prompted to select a custom database if you so desire – 
  5. Alright, final step. Here’s where we can change the port for user access (remember, this is what will be publicly facing) along with the certificate. I am staying with the defaults for my installation. 
  6. Now, we are ready to hit the Start Service button. Running…
  7. Success! 
  8. Now, we get a nice, simple portal that’s used to verify all of our services are operational 

Wow, a very streamlined process that was very intuitive while using a sleek and simple interface.

I am going to check to verify my org user can log into the portal…

Success! Our vCD authentication was passed through and now I get the vCAv C2C interface. 

Before I can have the paired sites and start testing workload migration, I need to go set up my SiteB vCloud Director instance and deploy the combined appliance. On the next blog post, we will go through the Site pairing process and do a migration/replication.

-Daniel

 

 

What’s VMware vCloud Availability for Cloud-to-Cloud DR?

As of yesterday (May 17th, 2018), VMware announced the release of VMware vCloud Availability for Cloud-to-Cloud DR 1.0 – release notes here while my esteemed colleague, Tom Fojta, announced this on Twitter:

Many of you may be wondering, what is vCloud Availability for Cloud-to-Cloud and how may I use this in the VMware Cloud Provider Program?

To start off, vCloud Availability for Cloud-to-Cloud (vCAv-C2C) is VMware’s solution to vCloud Director instance to instance disaster recovery and migration. Here’s a nice summary of what vCAv-C2C provides:

  1. Replicate and recover vApps (VMs) between two vCD instances for migration, DR, and planned migration use cases.
  2. Complete self-serviceability for the provider and tenant administrator. A unified HTML5 portal that will be utilized alongside vCloud Director. Replication, migration, and failover can be managed completely by the tenant or provided as a managed service by the provider.
  3. A simplified and streamlined architecture to support vCloud Director 8.20, 9.0, and 9.1 while supporting vSphere 6.0U3 and 6.5U1.

vCloud Availability for Cloud to Cloud DR Installation Documentation

In my opinion, point #3 is one of the most critical benefits to both providers and tenants. When we discuss multi-tenant architecture, this does tend to add layers of complexity, but the VMware Cloud Service Provider Business Unit has done a great job of rationalizing the architecture and streamlining it for vCAv-C2C and future solutions. I will get to the architecture shortly.

Before we get into the details of vCAv-C2C, many of you have experienced our other migration or disaster recovery-based solutions. I made this simple chart to showcase each of our current VMware Cloud Provider (CSPBU) solutions and how they complement one another:

As you can see, vCAv-C2C will complement the traditional vCAv solution while vCD Extender can still be used for on-prem tenant migrations to a vCD instance. vCAv-C2C fills a void on migration between vCD instances, which is a much-needed capability for our Providers.

So let’s talk about the high-level architecture. As I mentioned before, a lot of thought and development went into vCAv-C2C to make the architecture simplified and seamless. With vCAv-C2C, everything is packaged into a simple OVA deployment – no need to manually/CLI configure a vCAv deployment anymore. I was fortunate enough to be part of the alpha testing team (along with Fojta and my other peer Fernando Escobar) and was very pleased with this capability – ease of deployment and configuration is something that is required for many of our Providers.

Furthermore, this single OVA has every role required for vCAv-C2C. Per the documentation, we have a few roles:

  1. Replication Manager
  2. Replicator Node (Large Replicator role available too)
  3. Tunnel Node

Best of all, there’s a Combined role now that can be utilized for smaller or proof of concept (PoC) deployments. This is what I’ll be using in my lab environment.

Let’s talk about a high-level architecture –

As you can see, this is an appliance-based architecture that will protect (or migrate) vApps between site to site. Moreover, we can simplify this for PoC/small deployments by using a combined vCAv-C2C appliance –

Cloud to Cloud tunneling is utilized if you are going over a public internet connection and do not have private (VPN or Direct Connect) connectivity between the two vCD instances. VMware’s documentation writeup is here along with a nice drawing that depicts the DNAT and port requirements.

As for scale and concurrency guidelines, the team did a great job with support a significant amount of replications/migrations. From the release notes –

  • Scale Limits
    • 300 active protections for a single tenant
    • 300 active protections using a single large vCloud Availability Replicator. For more information about the replicator types, see Deploy vCloud Availability for Cloud-to-Cloud DR Services by Using the vSphere Web Client.
    • 1300 active protections across 20 tenants
    • 20 tenants with active replications
    • 7 active vCloud Availability Replicator instances
    • Up to 2 TB size of protected workloads
  • Concurrency Limits
    • 60 concurrent Protect, Test Failover, Reverse Protect, Test Failback, and Failback operations
    • 110 Concurrent Failover operations

If you’re a provider, you might be wondering how do I download the bits so I can start testing it?! Well, reach out to your respective VMware Cloud Provider field team as this is going to be an initial release and we want to work with our providers on ensuring all vCAv-C2C requirements are met for a successful deployment. You can also reach out to me directly and I’ll be happy to put you in touch with your respective team.

Up next – high-level installation instructions for vCAv-C2C!

-Daniel