- Integrated Systems
- About Us
- Integrated Systems
- About Us
vSphere 7.0 and HPE Nimble Storage
- vSphere 7 storage integrations with HPE Nimble storage including NCM 7.0 with vLCM
- vSphere 7, SRM 8.3 and vVols DR (focus of this blog)
- CNS support for vVols (published)
NimOS 220.127.116.11 and 5.2.1 which will release soon, will carry official support for vSphere 7.0 vcPlugin from HPE Nimble Storage. Watch the announcements from HPE Nimble Storage on the release.
vVols are not new to HPE Nimble Storage. From the technical beta in 2014 to the demos at VMworld 2016, Nimble Storage demonstrated DR capabilities with vVols VASA3.0 using PowerCLI. When the vVols replication DR feature was released with vSphere 6.5, both HPE 3PAR and HPE Nimble Storage had Day 0 support for this important feature required by many enterprise customers. The simplicity of the HPE Nimble Storage platform is something our customers appreciate. Here are the top 10 reasons to consider your move from VMFS or RDMs to vVols today.
The adoption of vVols continues to grow in 2020. In the last 6 months, we have seen our enterprise customers continue to push the scale, with 10,000 vVols supported on a list of our platforms. VMware shared vVol adoption numbers by vendor in Sept 2019:
vSphere Setup Pre-requisites
Since we want to setup SRM and disaster recovery, your setup should have 2 sites. Each site will have one vCenter, one SRM instance, and a few ESXi hosts. Before setting up SRM, there are some pre-requisites on your vSphere 7.0 environment to setup vVols. Do these sequentially as listed, on the vCenters and ESXi hosts on both sides. Screenshots are provided below for your reference, for each step.
- Setup NTP on your vCenter VCSA. Navigate to your vCenter VAMI at https://<vcenter-IP-or-FQDN>:5480/ and login as root. Under “Time”, enter a reachable Time server. You can also set timezone here, but that is optional.
- Setup NTP on each of your ESXi servers. Go to Configure --> System --> Time configuration --> Network Time Protocol --> Edit. Add your NTP server. Check the box to start the NTP service. Change the NTP service startup policy to “Start and stop with host” and hit OK. Once the change has been made, go back and check that it is indeed active. If you have a cluster with a large number of hosts, it is a good idea to use host profiles for consistent setup.
- If your HPE Nimble Storage array uses self-signed certificates, make sure ESXi can accept this certificate into its key store. See this VMware KB for additional details. Go to Configure --> System --> Advanced System Settings --> Check for a key Config.HostAgent.ssl.keyStore.allowSelfSigned, and if the value is “false”, select the entry and click “Edit”. The change to this setting does NOT require and ESXi reboot.
- This step is optional but helps debugging. Enable 2 services on ESXi: SSH and ESXi Shell. You can either edit the startup policy such that these services are always running when the host boots up, or simply start them when needed for debugging. Go to Configure --> System --> Services --> SSH or ESXi Shell and “Edit Startup Policy” or simply hit the “Start” button.
- For a Fibre Channel deployment, make sure your zones are configured for availability. We recommend failover redundancy, port 1 from FC card on ESXi zoned with controller A, and port 2 from FC card on ESXi zoned with controller B. Configure your ESXi networking to talk to the HPE Nimble Storage array, if this will be an iSCSI deployment. We recommend having 2 dedicated physical NICs on the ESXi on 2 separate L3 data networks. Additionally, a dedicated physical NIC for management and one for vMotion is highly recommended. Setting MTU to 9000 (Jumbo Frames) is recommended if you own your entire network and can run jumbo pings end to end. The HPE Nimble Storage array should also have the same two L3 data networks data1 and data2. The screenshots below show this recommended setup, and results of the jumbo ping to another ESXi for vMotion, and the array’s discovery IPs.
- vmk0: management vmknic on vSwitch0 with vmnic0 uplink
- vmk1: vMotion vmknic on vSwitch1 with vmnic1 uplink
- vmk2: data1 vmknic on vSwitch2 with vmnic2 uplink
- vmk3: data2 vmknic on vSwitch3 with vmnic3 uplink
- For iSCSI deployments, enable the software iSCSI adapter in ESXi. Go to Configure --> Storage --> Storage Adapters --> + Add Software Adapter --> Add software iSCSI adapter
If steps 1 through 6 seem like a lot of work, consider talking to your HPE Nimble Storage representative about our dHCI product line.
This is the environment I’ve setup:
- Site A
- vCenter: 10.21.123.210
- ESXi hosts: 10.21.170.159, 10.21.98.156
- Nimble array: sjc-array1099 with data discovery IPs 18.104.22.168 and 22.214.171.124
- SRM server: 10.21.26.24
- Site B
- vCenter: 10.21.28.179
- ESXi hosts: 10.21.28.185, 10.21.28.186
- Nimble array: sjc-array612 with data discovery IPs 126.96.36.199 and 188.8.131.52
- SRM server: 10.21.26.10
Setting up storage arrays for vVols and replication
Now it is time to setup the HPE Nimble storage arrays for each site. The array at site A will replicate to the array at site B, and vice-versa. To made administration easier, we will install and use the HPE Nimble Storage vCenter Plugin as well, on each site. Remember to perform these next steps on both sites.
- Register the vCenter with the Nimble array. Navigate to https://<Nimble-group-IP-or-FQDN> and head over to Administration --> VMware Integration --> Register a vCenter. Add an identifying name, and IP/FQDN of your site A vCenter host to this site A nimble array. Make sure do this on site B as well. Check both boxes for “Web Client” and “VASA Provider (VVols)”. This will push a new binary of the Nimble vCenter Plugin into the vCenter. You will see a brief message in a small ribbon on your Chrome browser about a plugin deployment, and a request to refresh your browser by simply clicking the button.
- Provider status must be “Online”
- Provider version must show the Nimble OS software version
- VASA API version must show “3.0”
- Setup the second array as a replication partner for the first, and vice-versa. Navigate to https://<Nimble-group-IP-or-FQDN> and head over to Manage --> Data Protection --> Replication Partners --> Create replication partner --> On-premises Replication Partner.
- Now let’s create a vVol datastore on both site A and site B. This is the datastore that will host VMs from site A, and when site A fails over to site B, the VMs will be in the datastore on site B. So, it is recommended to keep the name of this datastore the same on both sites. Follow step #2 from this blog to use the HPE Nimble Storage vCenter Plugin to create a vVol datastore. Confirm that the vVol datastore is mounted on all ESXi hosts, by navigating to “Storage” and selecting the vVol datastore. “Hosts” will show all hosts successfully mounting this datastore.
- Create VM Storage Profiles on both site A and site B. VM Storage Profiles help define the storage requirements of a VM. You can add capabilities exposed by your storage vendors that are required by the applications or the VMs. For this SRM demo, let’s say we want the data from VMs to be replicated once every hour. Go to vCenter --> Menu --> Policies and Profiles --> VM Storage Policies --> Create VM Storage Policy --> Name “srm-vvol-policy”. For “Policy Structure” select “Datastore specific rules” --> Enable rules for “Nimblestorage” storage.
Now, create a VM, and apply this policy to the VM. I'm choosing to migrate an existing VM to the vVol datastore. You could create, clone, clone from template, or migrate with the result of creating a new VM on this vVol datastore. While applying this policy for VMs that need DR protection, select a replication group (RG). Choose to create an RG of type “Automatic”.
When the VM create/migrate is complete, the summary page of the VM should show it is “Compliant” with the applied profile, and an RG is created automatically.
Note: selecting an “Automatic” RG at VM provisioning type has created an RG which maps to a Nimble volume collection. If we now create a second VM, and want it to be protected “together” with this VM i.e. on it’s exact same schedule, to behave like a cluster of app VMs together, while provisioning the second and subsequent VMs, the replication group drop-down will show an existing RG which can be selected.
At this point, if you were to use VASA3 DR API using PowerCLI, pyVmomi or any other SDK, you would have a fully functioning DR solution. This has been the working solution for the last 4 years or so.
Next, we are ready to setup SRM 8.3 that includes additional UI to facilitate vVol VM DR. Watch this space for the next blog.
- haniff on: High-performance, low-latency networks for edge an...
- StorageExperts on: Configure vSphere Metro Storage Cluster with HPE N...
- haniff on: Need for speed and efficiency from high performanc...
- haniff on: Efficient networking for HPE’s Alletra cloud-nativ...
- CalvinZito on: What’s new in HPE SimpliVity 4.1.0
- MichaelMattsson on: HPE CSI Driver for Kubernetes v1.4.0 with expanded...
- StorageExperts on: HPE Nimble Storage dHCI Intelligent 1-Click Update...
- ORielly on: Power Loss at the Edge? Protect Your Data with New...
- on: HPE Primera Storage celebrates one year!
- on: Introducing Language Bindings for HPE SimpliVity R...