Datacenter Infrastructure & Disaster Recovery Plan

/Datacenter Infrastructure & Disaster Recovery Plan
Datacenter Infrastructure & Disaster Recovery Plan2018-09-19T07:55:16+00:00

General Infrastructure:

  • Dedicated scalable servers start with the following:
    • All Cisco/DELL Enterprise Hardware (Dell R610 – 2 x Intel® Xeon® E5645 2.40GHz CPUs, 16GB 1333MHz Ram, Raid 10 w/ Enterprise SSD Storage)
    • Windows Server 2012 R2 or CentOS 7 Linux Operating System
    • LiteSpeed Web Server (faster than Apache and Nginx)
    • Percona (MySQL) Database Server
    • Support for PHP 5.6+, Prototype/jQuery/AJAX Development Languages
    • EdgeCast/MaxCDN Content Delivery Network (1,500 GB CDN Data Usage /month)
  • Multi-layer backup protection:
    • Daily backups in the datacenter (One Wilshire Downtown Los Angeles), then uploaded to multiple locations in the cloud delivering offsite backup for maximum failover protection.
    • If datacenter was to go down, we would have a backup of the server in the cloud we would access and recover in another datacenter quickly.
  • Security & Firewall Protection:
    • Datacenter Protection:
      • Certification: SSAE 16 Type II
      • Biometrics/security card access and DVR camera systems
      • The Network Operations Center (NOC) staff monitors the network 24x7x365
      • Environmental Control: State of the art climate control systems
      • Fire Suppression: Pre-action dry pipe sprinkler system with clean agent fire extinguishers
    • Hardware AND Software Firewall: Cisco ASA Firewall Appliance before Servers, ConfigServer Linux Software Firewall on Servers
    • ModSecurity HTTP transaction logging & exploit protection
  • System Availability:
  • Bandwidth:
    • Determined by expected traffic and historic analytics reports. We can turn up the bandwidth and scale server power as needed, as we offer a dynamic solution.
  • Email Functionality:
    • Application Server will also handle e-mail transacting, or we can use an external SMTP Auth server.
  • Interoperability:
    • SSH access (via whitelisted IP address) will be provided so that your developers and database administrators can work within the production environment.
  • Supportability:
    • A dedicated 8/5 support representative and 24/7 emergency support is available to you and your client.
    • We staff certified developers, so any coding assistance can be utilized at $95/hour.
    • System can be upgraded at any time with little downtime, as mentioned, our private cloud is very dynamic.
    • Logging is turned on both the Web application server and the database server in order to monitor any issues with your environment.
  • Scalability:
    • If additional memory or CPU power is ever needed, we can scale up with only seconds of downtime. We may also utilize an enterprise Content Delivery Network (CDN) to help offset server load and mitigate DDoS attacks.

Disaster Recovery Plan (2 Parts):

1)     Near Zero-Downtime Protection:

  1.      Automatic DNS Failover: 6 DNS servers for maximum failover and availability.
  2.      Disk Redundancy: All hard disks are in Raid mirrored configuration and can be hot swapped and replaced with ZERO downtime.
  3.      Onsite Backups: We take daily snapshots of all databases, and weekly snapshots of all files, this frequency can be increased for an additional fee.
  4.      Offsite Backups: We store ALL backups in our cloud, over multiple locations throughout the United States to geographically keep data safe, secure, and highly available.

2)      Recovery Plan:

Due to high redundancy and availability of network and data, the only way we would incur a serious problem is if a major disaster were to hit downtown Los Angeles. While this is highly unlikely, our plan to counter such occurrence would be the following:

  1.     If One Wilshire Los Angeles is affected by a major disaster, we will replicate the client’s server instance (cloud snapshot) in one of our datacenters closest by. We have a 10GB fiber network connecting several datacenters across the globe (Estimated downtime 4-7 hours depending on amount of data to recover):
  2.     United States: Los Angeles, Seattle, San Jose, Miami, New York
  3.     International: Toronto CA, Vancouver CA, Portsmouth UK
  4.     Another option is to have multiple dedicated servers in two geographical locations connected in our private cloud, which would create 100% uptime in the unlikely event something were to happen in Los Angeles. However, the cost for hosting will double in price as we’ll need twice the hardware and bandwidth. It is important to note that while operating a completely redundant secondary disaster recovery site can offer the fastest service restoration times, it is the most costly alternative, effectively doubling the cost of operations.