Capax Global offers best-in-class enterprise hosting. Our datacenters are capable of supporting the most robust and scalable infrastructure
implementations available.
Show All |
Hide All
24x7x365 - Our data centers are continuously operational and always staffed with personnel capable of handling planned and unplanned
maintenance operations
Three 9s - Our uptime is currently measured at 99.9% which equates to 8 hours downtime per year
Monitoring - Of key system health indicators (disk space, CPU/Memory utilization, etc)
Alerting - Threshold based alerting on key system health indicators
Firewall - Redundant and dedicated external firewalls limit external traffic to only the essential traffic
Segmentation - Non-related systems are segregated on different network segments. The resulting compartmentalization
results in higher security
Remote Access - Remote access is provided to specific systems via VPN
Hardening - Operating systems are configured with the minimal possible exposure to external threats
High Speed - All internal network components run at 1Gbps. Our external network connection is 45Mbps burstable to 100Mbps
or higher as needed
Switching - All internal networking is switched using virtual segmentation
Load Balancing - Network hardware load balancers are employed for increased scalability of application and search services. All load balancer
interface ports speeds are 1Gbps
Throttling - External connection load can be throttled in order to prevent denial-of-service attacks or system overload during heavy load periods
DNS - Standard naming service support
DHCP - Standard dynamic IP address support
Network - The internal network is 1Gbps with upgrades to higher speeds as stable implementations become available. The external 45Mbps
redundant connection to the Internet is burstable with additional bandwidth available on a contractual basis
Storage - Storage scalability is provided by the mirroring systems. Each system provides additional internal storage in order to hold more
data. External SAN storage is also available on a contractual basis
Backup - Backup and restore services are available on a contract basis. Backups may be made to online storage (disk) for quick restoration or
to offline (tape) storage. Offsite tape archival is also available upon request
Databases - Capax Global can host and support mySQL and SQL Server 2005 databases. Additional database support can be added on a contractual basis
Availability, also referred to as reliability or up-time, is the measure of how well your system limits downtime. Capax Global currently
supports 99.9% availability. That means the systems are available 99.9% of the time. We support this high degree of availability through
redundancy at all levels as follows:
Multiple Sites - Our redundancy extends to the physical data center. We currently have two data centers, one on each coast. When combined with the
additional redundancy features listed below, this allows our services to survive major disasters, be they natural or man-made
Power - The data center receives power from at least two separate sources. The power lines follow diverse routes into the building. The data
center has one alternate generator capable of powering the data center indefinitely. The data center employs two separate UPS systems
both with the capacity to power the entire data center for long enough period to bring the generator online. UPS failover is instant
and undetectable by datacenter systems. Every hardware system has at least two power supplies. Each power supply is powered through a
separate UPS system with diverse power line routing
Network - The data center receives network services from all top tier service providers for external network access. The external network
connections follow diverse routes into the building. Every piece of network equipment is redundant. Each computer system has at least
two network interfaces. Those network interfaces are connected through separate network switches with diverse network cabling
Load Balancers - Load balancers are deployed in redundant pairs (i.e. clusters) configured for instant failover. Each load balancer is connected
to both switches with diverse routing of network cabling
Disk - The Short Term Partition will use the local disk mirror. Due to their redundancy, the Long Term Partition indices are not strictly
required to have disk redundancy, however hardware disk mirroring is employed. The Archive Partitions mirror sets are smaller and
therefore more exposed to data loss, so we employ at least hardware mirroring for our disk systems
Indices - It is possible, and sometimes desirable, to host more than one index on a single host computer system. Such configurations will be
tuned based upon the specific load requirements of the system. Any given mirror set is comprised of at least two computer host systems
in order to survive a single host system failure. Each index has its own set of physical disks unless performance requirements are so
low that disks can be shared
DAH Query Dispatchers - It is possible, and sometimes desirable, to host more than one Query Dispatcher on a single host computer system. Such configurations
will be tuned based upon the specific load requirements of the system. Any given load-balanced set of Query Dispatchers is comprised
of at least two computer host systems in order to survive a single host system failure
DIH Index Dispatchers - It is possible, and sometimes desirable, to host more than one Index Dispatcher on a single host computer system. Such configurations
will be tuned based upon the specific load requirements of the system. Any given load-balanced set of Index Dispatchers will be
comprised of at least two computer host systems in order to survive a single host system failure