MANAGED STORAGE SERVICES
Logicworks can design and implement storage solutions that range from small, high-performance systems to 10s or even 100s of Terabytes. Since moving large amounts of data over the network, even a storage area network, can take a significant amount of time, Logicworks takes special care in designing solutions to ensure that clients can either failover immediately or rapidly recover in the event of storage device failure.
Logicworks offers servers with 8-bay internal SAS and/or SATA storage and 16-bay internal SCSI storage, which provide a viable alternative to Direct Attached Storage (DAS). Internal storage is ideal for Distributed Replicated Block Device (DRBD) based solutions, providing the most cost-effective high-availability storage solution available.
When internal storage isn’t sufficient, external Fibre Channel-attached SAS and SATA storage arrays provide ample additional storage capacity. Storage arrays are available either in 12- or 16-bay configurations, are based on 4Gb/s Fibre channel, and allow for JBOD arrays to be daisy chained off the primary controller for up to 80 disks of direct attached storage. Logicworks configures arrays in either RAID 1+0, 5, or 6 levels, along with at least one hot spare, ensuring minimal loss of fault tolerance, even during RAID rebuilds.
Logicworks provides Fibre Channel switches for Storage Area Networks (SANs) on a dedicated basis, ensuring that no other client’s I/O affects another’s. SANs enable shared storage required by database clustering, and provide the greatest possible flexibility in storage management. As with Ethernet switches, Fibre Channel switches are typically deployed in pairs, ensuring no single point of failure. Scaling for anticipated or unanticipated growth with a SAN solution is fast, easy, and cost-effective.
DRBD is a high-availability storage solution ideal for shared filesystems such as NFS, MySQL, and other Linux-based applications. DRBD solutions require a pair of servers, either with on-board or direct attached storage. The primary server replicates all writes to the DRBD block device over a network connection to the secondary server. In the event the primary fails, the secondary server remounts the block device with write access, ups the VIP and starts the relevant service. Since replication takes place over the network, the secondary DRBD node can be located in an alternate facility. DRBD is a proven, highly resilient, high-performance storage solution that provides high-availability in a “shared-nothing” configuration.
Since any high-availability system or replication configuration will also replicate corrupt data and deletions, all solutions require backup and recovery for ‘point-in-time’ recovery. Logicworks offers both Continuous Data Protection (CDP) and file-level backup services. Backing up is only half (or less) of the battle – recovery, and the time it takes to do so, is the critical performance metric of any backup solution. All Logicworks backup solutions are written to two layers of disk -- the first local, and the second offsite – before being written to tape. Tape, in turn, is shipped to Iron Mountain and is only intended to be used for recovery from a severe disaster. Since restores a read from disk, there is no waiting for tape to spool to the right location, and no possibility of media failure, so your data is back to its desired state as fast as it can be copied over the network.
