Storage Platforms
A key component for organization infrastructure deployment is a secure and redundant storage design offering protection for corporate data against threats and hardware failure. Advanced storage solutions must be designed to work within existing redundancy and high security models while providing high performance connectivity to local and remote clients. Storage platform design will incorporate redundancy protocols and encryption services.
Altexxa Group works primarily with ZFS driven storage solutions and has had great success with FreeBSD based storage design. We believe the deployment of redundant storage cluster systems with data replication solutions will provide the required level of protection for critical application data.
- Protection for corporate data against threats and hardware failure
- Working within existing redundancy and high security models
- Integrated encryption services for physical layer protection
- Clustered design with data replication for critical applications
Storage Design
Altexxa Group works with different aspects of storage design and implementation based upon the application performance and capacity requirements. Our direct-attached storage design includes SAS/SSD performance hardware with fiber channel connectivity and is used for small application deployment requiring high performance local storage. Working with a hypervisor node cluster design, storage requirements include the use of a separate storage network featuring FCoE or iSCSI connectivity. We also work with a tiered storage categorization offering specific performance and redundancy metrics allowing a storage implementation to meet certain application needs. Additional considerations for storage design will include compatibility with a DR model and use of data replication between the primary site and remote hot-site.
- DAS and SAN design and implementation based on application requirements
- Clustered storage solutions for working within a virtualization model
- Tiered storage levels for meeting specific usage demands
- Data replication considerations for Disaster Recovery implementation
Connectivity
Following the standard redundancy model, storage connectivity across a network-attached design will utilize multiple paths and controllers. Our use of fiber channel solutions involves the deployment of at least two separate paths between the host nodes and the storage chassis. The redundancy model requires the use of separate node HBAs, FC switches and SAN controllers. Connectivity for Ethernet driven storage systems follows the same requirements and will feature protocols such as LACP within the deployment.
Additional considerations for storage connectivity include replication needs and active/standby configurations. Altexxa Group believes in following established redundancy models through the implementation of a multiple path network layout and transport protocols.
- Following redundancy model by utilizing a multiple path design
- Node and clustered storage connectvitiy driven by Ethernet and Fiber Channel
- Physical and logical network considerations for establishing redundancy
- Evaluating capacity requirements for deployment considerations
Deployment
A truly scalable storage deployment should allow for cluster expansion in stages to match growing capacity requirements. Critical storage infrastructure such as the network backend should utilize 10/40Gbps Ethernet or 16Gbps FC connectivity with ability to expand the fabric. Following a cloud deployment model for distributed storage is an important factor for working with disaster risk zones and following data redundancy requirements to achieve logical data protection.
The Altexxa Group storage deployment method also accounts for data backup requirements both at block and file level. We rely on automation systems to balance loads across storage clusters for improved latency. We also analyze OS-level data storage requirements to optimize file system usage.
- Scalable deployment to allow for projected growth
- Backed by fast Ethernet or Fiber Channel connectivity
- Distributed design for optimal data protection
- Data backup and automation considerations