What is CloudSimple and what do they do? Why did Google acquire them? What does it mean for enterprises? What about VMware data protection in Google Cloud? These are some of the questions that I will address in this blog.
What is CloudSimple
CloudSimple runs and manages VMware stack on bare metal compute in Microsoft Azure and Google Cloud Platform (GCP) making it simple for enterprises to run VMWare in Azure and GCP. Enterprises can subscribe to the desired amount of compute, memory, storage in the cloud and run VMware natively in the cloud. Enterprises pay per consumption and can grow and shrink their VMware real estate in the cloud on-demand.
CloudSimple Benefits for Enterprises
Every enterprise wants to reduce their data center footprint and leverage cloud to address on-demand services at scale, and pay-per-use economics. Enterprises also prefer keeping their operational model consistent across on-premises and in-cloud environments. CloudSimple solution in GCP’s bare metal cloud offers several advantages: :
- Continue to operate the familiar on-premises VMWare ecosystem in GCP: vCenter, NSX, vSAN …
- The proximity of apps in GCP enables them to leverage other native GCP offerings such as Google BigQuery for Analytics, Google Nearline/Coldline object storage to store backups, replicate to a remote GCP region for DR.
- Self-service, on-demand consumption model for Dev, QA, UAT, DevOps teams which accelerate application test and release cycles.
- Enterprise apps such as SAP and other custom apps etc. will continue to run as-is with their current licensing models as-is in VMware in GCP.
CloudSimple Benefits of VMware in Google Cloud as compared to VMware in AWS?
AWS and VMware worked together to come up with VMC (VMware Cloud) in AWS. One of the big challenges with this solution is the lack of critical capabilities in the vCenter/ESX environment, which operations teams are very familiar with and deeply invested in.
CloudSimple in GCP makes it really “simple”. Operations continue to get all the capabilities of vCenter/ESX environment so they can continue to run all their existing applications, backups, DR run book etc as-is with almost no changes in their day-to-day operations.
Data Management for VMware in Google Cloud Platform
Enterprises never consider running production applications in a platform until they are satisfied that there is a robust, enterprise-grade, security-hardened backup and DR solution.
With access to vCenter, hypervisor, and VADP APIs, most backup and DR solutions should work as-is. i.e. they should be able to discover VMs, leverage VADP APIs to protect the VMs, instant-mount the VM backups over NFS or iSCSI and spin up the VMs.
But, the lift and shift approach to the cloud comes with its risks … especially the very high costs of data management in the cloud. Thus, while evaluating a data management solution for VMware in GCP, you should look for the following critical capabilities.
1. Backup to cloud object storage: Storing backups in vSAN storage in the cloud is very expensive. So, consider a solution that doesn’t need to store backups in expensive block storage ($170/TB/month). Instead, it should store VMware backups directly in object storage such as Google Nearline ($10/TB/month), Google Coldline ($7/TB/month).
2. Instant Mount & Recovery from Google Cloud Object Storage: Most backup products need to rehydrate backups from object to block storage before providing access to VMs. This increases the RTO.
To reduce RTO, some vendors auto-restore backups from object to bock storage in the background. This, unfortunately, increases the TCO because of 24×7 bock storage costs (Persistent Disk costs $170/TB/month) in GCP.
Look for a solution that can mount backups from object storage and recover VMware VMs instantly. And while the VMs are up and running, it needs to cache reads and writes in bock storage thus delivering up to 80% of the block storage performance at just 20% of the costs. And of course, in the event of a real disaster of a mission-critical DB, you should have the option to storage v-motion to the desired storage-tier after the instant mount and recovery.
3. Push-button DR orchestration from on-premises to VMware in GCP: Enterprises should be able to create DR plans, thus allowing which VMs to recover in what order, in what network, with how much CPU and memory, and then execute one or more DR plans with the push of a button. This reduces the operational burden at the time of DR or DR tests when operations are under severe pressure to demonstrate a successful DR.
4. Copy Data Management for applications in VMware in GCP:
a. Copy Data Management is the management of all copies of production applications i.e. backups, disaster recovery, database cloning.
b. A lot of enterprises use containers in test/dev environments. Look for a backup vendor which has the capability to reuse the backups and provision storage space efficient rapid virtual database clones for SAP, MS SQL, Db2 and others to containers running in GCP.
c. Many enterprises want to use on-demand data warehousing and analytics in the cloud. Look for a backup vendor that stores application backups in native application format and enables you to mount the backups instantly so the data can be fed to Google BigQuery for on-demand analytics and data warehousing.
5. Native Google Cloud Marketplace offering: Look out for a SaaS offering in GCP marketplace that can support backup, DR, migration, DB cloning for workloads running anywhere. This enables you to consume Copy Data Management i.e. Backup, DR, DB Cloning in a subscription model with integrated billing from Google. It also enables your operations to have a single pane of glass for data management for workloads running on-premises, in VMware on GCP, and in GCP native workloads.