Actifio Sky for Google Cloud
Protect on-prem workloads as well as workloads running in Google Compute Engine (GCE) and replicate to another Google Cloud region for on-demand disaster recovery, long term data retention, and on-demand database cloning.
Recover Multi-Terabyte Data in Minutes (SAP HANA, MS SQL, and more)
Actifio Sky running in Google Compute Engine (GCE) uses persistent disk to store backup data. It incorporates powerful data virtualization technology to instantly mount backup data to on-demand GCE instances and delivers low Recovery Time Objective (RTO).
Sky integrates with applications and databases such as SAP HANA, Microsoft SQL Server in Windows and other databases in Linux to deliver instant read/write enabled data volumes for instant recovery purposes.
Compare the low RTO, in minutes, delivered by Actifio Sky with other backup products which can provide hours or days at best.
Satisfy Long Term Data Retention Needs
For data that must be retained for days or months, Actifio Sky can use google persistent disk with HDD and/or SSD. For data that needs to be retained for years or decades Actifio Sky uses Google Cloud Object Storage such as Google Nearline.
Actifio Sky incorporates the unique ability to deliver instant access to data stored in Google Nearline Object Storage irrespective of whether the data is days or decades old. Compare this capability with other backup products that can take hours or days to rehydrate deduplicated data from object storage.
Secure Data with Encryption and Satisfy Compliance Needs
All data is encrypted using AES256 when replicated to remote Actifio Sky instances in GCE. GCE encrypts data before writing to persistent disks. You have the option to specify encryption keys. Actifio also provides role based access control and an audit trail for added security. To satisfy your compliance needs, Google maintains the following certifications – ISO 27001, SSAE-16, SOC 1, SOC 2, SOC 3.
Clone Multi-Terabyte Databases in Minutes for DevOps
Information that is protected by Actifio Sky in persistent disk for backups can also be leveraged to provision multiple virtual database clones for Dev, QA, UAT, Analytics or other test environments. Sky also enables developer self-service to simplify and accelerate the provisioning of virtual database clones.
Re-writable virtual clones are space efficient and do not use any extra storage. Incremental disk capacity is only consumed when the database is changed as part of the testing process.
Other backup products typically take hours or days to create physical clones with 10x or more storage required in test environments.
Reduce Backup Window
After the first full backup, Actifio performs incremental forever backups for everything i.e. SAP HANA, MS SQL, PostgreSQL & other databases, VMs, physical servers and NAS filers. This process reduces the backup window by up to 10x and significantly decreases the storage IO and network IO of production applications.
Actifio’s incremental forever model dramatically accelerates the protection of multi TB databases and files systems with millions of files. Compare this strategy with traditional protection approaches that rely on weekly full and daily incremental backups which lead to long backup window and significantly impact production applications.
SLA-Driven, Application Consistent Backups
Actifio Sky includes deep integration with various databases such as MS SQL, PostgreSQL, MySQL, and applications such as SAP HANA, MS Exchange and Sharepoint. This deep integration ensures application consistent backups. For example – Actifio Sky uses SAP HANA APIs, Oracle RMAN for Oracle databases, Microsoft VSS snapshot for Windows based applications.
For large and geographically distributed environments, Actifio Global Manager (AGM) provides a single pane of glass to manage systems. AGM allows you to setup backup, replication and retention SLAs for applications in multiple data centers and cloud regions.
Actifio Sky Uses the Following Various Components of Google Cloud
- Actifio Sky can run in the following Google Compute Engine models: n1-standard-2, n1-standard-4, n1-standard-8, n1-highmem-8, n1-highmem-16
- Actifio Sky can use standard and SSD persistent disks to store backup images.
- Actifio Sky can use object storage such as Google Cloud Storage.
Actifio Allows for Simple Data Vaulting to Google Cloud Storage
Actifio lets you capture, manage, and use application data more quickly, efficiently, and simply than ever, across a whole range of business resiliency and agility use cases. Create simple, intuitive SLA’s to determine which data will be managed by the system, how often it will be captured from production, where it will be kept, and for how long. Then apply any SLA to any application, with just a few clicks.
Now Actifio customers can just as easily add a high performance “Vault” profile to their SLA portfolio, enabling them to move application data directly into Google’s new long-term storage platform, Google Cloud Storage. With just a click, you’ll get long term storage economics (~1¢/GB for data at rest) with immediate availability performance (~3 second response times for data retrieval.)
Frequently Asked Questions
If data is encrypted and stored in Google Cloud Storage, how will does decryption work?
Do users have to have their own Google accounts to take advantage of this service?
Will Actifio store data in Google Cloud in deduped format?
While Actifio stores data in deduped format when using block devices, in the case of Google Cloud Storage, data will not be stored deduped. Two key reasons are:
1) Storage economics have changed. Dedup was invented when storage was expensive. Google Cloud Storage costs $10 per TB per month, which is extremely inexpensive with almost zero management costs.
2) More importantly the value derived by storing data un-deduped in Google Cloud Storage is that the data can actually be accessed very quickly for restores, etc. We believe that the real value of storing data in Google Cloud Storage (aside from compliance reasons) is in accessing the data for analytics and mining purposes. Adding dedup adds overhead when users want to mine/analyze the wealth of data they have stored in Nearline.
Data can be accessed very quickly from Nearline, but does data have to be read OUT of Google Cloud Storage and stored in some local block storage in order for users to access the data? For example: If a user needs to mine data that existed in a 10 TB volume from last year, will they have to read all the 10 TB from Google Cloud Storage to a temporary 10 TB volume in their data center in order to start analyzing the data?
In the aforementioned example, Actifio doesn’t have to restore the 10 TB data from Google Cloud Storage to a data center. Because Actifio virtualizes data, all users have to do with Actifio is using the Actifio UI, mount the 10 TB volume straight from Google Cloud Storage as a mount point. This can happen in as little as a minute, and then the user will see the entire file system as if it exists in their own data center. Afterwards users can access whatever files or blocks of data they want. This is what we refer to as instant access.
Apart from instant access to data in Google Cloud Storage, Actifio also reduces bandwidth spend by keeping data extraction OUT from Google Nearline to a minimum. Therefore, if a user wants to access a 10 GB file out of a 10 TB volume, they don’t have to copy the entire 10 TB. They just access the 10 GB file and then get charged by Google only for that 10 GB file.
If I have workloads in AWS, will I be able to use Google Cloud Storage? For example, I have enterprise applications hosted in AWS, including MS SQL and Oracle databases, along with a few other VMs which have unstructured data.