The additional information that is needed to define the implementation process later is how much backup data is stored on the old hardware. This information is important for designing and sizing the migration strategy and timeline for moving the backups from the old hardware to the new hardware. For example, you can use the amount of backup data to estimate how long it will take to copy or move the backups to the new storage devices. You can also use the amount of backup data to determine whether you need to use compression, deduplication, or WAN acceleration to optimize the migration traffic.
References: [Migrating Backup Files], [WAN Acceleration]
Topic 3, Veeam Financial Services
Solution Overview
Veeam Financial Services will replace the current backup solution with Veeam Backup & Replication. They plan to protect their virtual infrastructure in New York, San Francisco, London and Singapore. Each branch houses a single physical Windows server that will also need to be backed up. With their stance towards the cloud, Veeam Financial Services would also benefit from any cloud integration where possible. A user portal is also being developed that will enable limited self-service restores for its user base.
Business requirements
Any backup solution should leverage public cloud storage. Mission-critical applications should not be affected by backup processes. To reduce the workload on the backup management team, every possible way to automate the solution should be used. To meet regulatory requirements, backups must be retained for four years. Veeam Financial Services has plans to develop a self-service portal and integration with the new backup solution is required. Ticket creation in their current help desk system must be automated based on alerts generated. Local privacy and data locality laws must befollowed for each branch. Due to the confidential nature of internal financial data, all security features available in backup software should be utilized. Backups should be dynamically scoped to reduce administrative overhead. The backup solution should not require proprietary hardware or storage.
Existing technical environment
o Each site has multiple VMware clusters, including a management cluster, general purpose clusters, application clusters, and Linux/Oracle clusters.
o Each site has a virtualized stock trading application that has extremely high disk I/O that is very latency sensitive.
o A custom financial application resides on AIX servers at each site.
o There is a 2 Gbps link between all major sites.
o Each branch office has a 100 Mbps connection to the nearest main site.
o Clustered NAS devices at each primary site house financial transaction and regulatory information, as well as the company's document repository.
o Each NAS cluster has an average of 100 shares, consisting of NFS 4.1 and SMB 3.0 shares.
o All traffic routed between LANs must pass through a firewall.
o Local network bandwidth at each site is 10 Gbps.
o Each primary site has multiple SQL instances, to include physical, virtual and always-on clusters.
o Each site has multiple physical and virtual Oracle servers.
o Each region has a separate Active Directory with no trusts established in any other region to meet regulatory requirements.
o Current backup storage is proprietary deduplication appliances that only support the current vendor's software.
o Each site has 10 production VLANs, one management LAN and three VLANs for physical servers.
o In North America, there is one VLAN that spans between New York and San Francisco.
o All virtual machines reside on a combination of SAN storage and NFS datastores.
o Each location has multiple physical Windows servers that act as file servers.
o VMware clusters can communicate, but all traffic between them must pass through a firewall.
Technical requirements
o All backups should be completed between 7 p.m. and 7 a.m. local time.
o All workloads should be dynamically grouped for ease of administration when creating and maintaining backup jobs.
o A copy of the backup data should reside in a local public cloud for each region.
o All backup repositories should be hardened.
o Any space-saving capabilities within the backup product should be leveraged.
o Any data transporter should scale horizontally rather than incrementally.
o The disaster recovery capabilities of the solution shall be utilized in the North American region between New York and San Francisco.
o Any storage integration shall be implemented in software.
o An immutable copy of all backups must be stored off-site.
o Twelve weekly backups, six monthly backups, and seven annual backups must be stored for archival purposes.
o Oracle backups must be native to Veeam software, with no Oracle administrative plug-ins.
o All data must be encrypted in flight and at rest.
o There must be three copies of data, on two different types of media, with one copy off-site that is unalterable or air-gapped.
o Only non-domain accounts are allowed for vendor components.
o Depending on regulatory requirements, some data may need to be deleted from backups before a restore can be performed.
o All backups must be scanned for malware prior to performing a restore.
o Gold level machines have a recovery time target of 30 minutes and a recovery point target of one hour.
o Silver level machines have a recovery time objective of six hours and a recovery point objective of 24 hours.
o Bronze level machines, which consist of development machines and field portables, have a recovery point target of seven days, with no recovery time target set.
o On-site storage must accommodate three years of projected capacity.
o All gold and silver machines require backups to be verified.
o All laptop backups must be managed completely without user interaction.
o Alternate backup methods should be used for virtual machines that are mission critical and susceptible to stuns