The ultimate solution for building reliable high-performance storage systems for mission-critical database applications in telecommunications requires a combination of industry standards: the CompactPCI system bus architecture; the RAID (Redundant Array of Independent Disks) mass-storage methodology; and the SCSI peripheral bus.
Besides providing high performance and excellent reliability, availability and serviceability, these standards offer the built-in flexibility and versatility for building systems that meet a broad range of different price/performance goals. Moreover, they are future-proof, constantly improving their ability to deliver fault-tolerant storage and maximum I/O performance for high-performance embedded systems.
For RAID systems and many other applications, the passive-backplane CompactPCI bus provides a more reliable alternative to the motherboard-based PCI bus and a more versatile and inexpensive alternative to VMEbus. In contrast to the edge-card connectors used in PCI, Compact PCI utilizes a high-reliability pin and socket connector system with shielding for EMI/RFI protection. In contrast to VMEbus, with its limited-capacity DIN connector, CompactPCI utilizes a 2- millimeter connector with a high pin count for I/O to be conveniently routed to the rear of chassis, as well as providing a vehicle for auxiliary buses or other, user-defined functions. Rear I/O results in a clean, uncluttered front panel, and it simplifies servicing of the system.
In an increasing number of applications where lost data and downtime are unacceptable, RAID has become the technique of choice for managing mass storage. The use of RAID algorithms in managing mass storage can improve both data accessibility and data reliability individually or simultaneously. There are numerous levels that can be implemented as well as competing system bus technologies and dueling peripheral bus technologies. At this point in time, a combination of the CompactPCI system bus and SCSI peripheral bus provides the best combination for the widest range of fault tolerant data storage applications.
Each RAID level makes use of one or more basic methodologies: striping, which distributes data among the disks in an array; mirroring, which stores identical data on multiple drives; and parity, which first calculates the data necessary to restore the contents on a drive if it fails, and then stores that data on a different disk drive. Each RAID level has its associated tradeoffs in performance, protection and cost and, depending on the systems I/O throughput performance requirements and the fault tolerance needs of the data stored, will determine which RAID level is appropriate.
A good case in point is the use of an embedded CompactPCI/RAID controller application in the Service Control Point (SCP) function in an SS7 network central server The SCP supports the database storage functions required in a SS7 network. Typically, the SCP receives database storage and retrieval requests from the Signal Transfer Point (STP), processes those requests and then either stores or retrieves and forwards the appropriate information to the STP. Due to their mission-critical nature, SCPs are deployed as redundant pairs although not necessarily next to one another.
Utilizing CompactPCI architecture, each SCP can be configured with a host CPU board, two WAN I/O boards and an embedded RAID controller board all connected with each other across a CompactPCI bus. For database storage and retrieval request servicing, each SCP would communicate with its corresponding STPs utilizing appropriate WAN links. The RAID controller contains the Ultra2 SCSI channels needed to interface with the disk arrays. A disk array of up to 15 individually accessed disks, can be attached the each Ultra2 SCSI channel located on the RAID controller.
Because SS7 database files are relatively small and read/writes are quite frequently, RAID Level 5,with disk stripping with distributed parity which generally perform well with I/O request intensive loads, would be the appropriate configuration for each SCP. RAID Level 5 has excellent read performance, but its write performance suffers somewhat since it takes time to calculate and then write parity data. Caching and other techniques can be used to improve write performance. RAID Level 5 also provides a higher degree of availability over levels using a dedicated parity drive, which represents a single point of failure. If a drive fails, its contents can be reconstructed from the remaining drives.
See related chart