Business Continuity and Disaster Recovery¶

2020.1

The Changineers Contingency Plan establishes procedures to recover Changineers following a disruption resulting from a disaster. This Disaster Recovery Policy is maintained by the Changineers Security Officer and Privacy Officer.

Policy Statements¶

Changineers policy requires that:

(a) A plan and process for business continuity and disaster recovery (BCDR), including the backup and recovery of systems and data, must be defined and documented.

(b) BCDR shall be simulated and tested at least once a year. Metrics shall be measured and identified recovery enhancements shall be filed to improve the BCDR process.

(c) Security controls and requirements must be maintained during all BCDR activities.

Controls and Procedures¶

BCDR Objectives and Roles¶

Objectives¶

The following objectives have been established for this plan:

Maximize the effectiveness of contingency operations through an established plan that consists of the following phases:
- Notification/Activation phase to detect and assess damage and to activate the plan;
- Recovery phase to restore temporary IT operations and recover damage done to the original system;
- Reconstitution phase to restore IT system processing capabilities to normal operations.
Identify the activities, resources, and procedures needed to carry out Changineers processing requirements during prolonged interruptions to normal operations.
Identify and define the impact of interruptions to Changineers systems.
Assign responsibilities to designated personnel and provide guidance for recovering Changineers during prolonged periods of interruption to normal operations.
Ensure coordination with other Changineers staff who will participate in the contingency planning strategies.
Ensure coordination with external points of contact and vendors who will participate in the contingency planning strategies.

Example of the types of disasters that would initiate this plan are natural disaster, political disturbances, man made disaster, external human threats, and internal malicious activities.

Changineers defined two categories of systems from a disaster recovery perspective.

Critical Systems. These systems host production application servers/services and database servers/services or are required for functioning of systems that host production applications and data. These systems, if unavailable, affect the integrity of data and must be restored, or have a process begun to restore them, immediately upon becoming unavailable.
Non-critical Systems. These are all systems not considered critical by definition above. These systems, while they may affect the performance and overall security of critical systems, do not prevent Critical systems from functioning and being accessed appropriately. These systems are restored at a lower priority than critical systems.

Line of Succession¶

The following order of succession to ensure that decision-making authority for the Changineers Contingency Plan is uninterrupted. The Chief Operating Officer (COO) is responsible for ensuring the safety of personnel and the execution of procedures documented within this Changineers Contingency Plan. The Head of Engineering is responsible for the recovery of Changineers technical environments. If the COO or Head of Engineering is unable to function as the overall authority or chooses to delegate this responsibility to a successor, the CEO shall function as that authority or choose an alternative delegate. To provide contact initiation should the contingency plan need to be initiated, please use the contact list below.

Sonya Corcoran, COO: info@changineers.com.au
James Gregory, Head of Engineering: info@changineers.com.au
Renee Lim, CEO: info@changineers.com.au

Response Teams and Responsibilities¶

The following teams have been developed and trained to respond to a contingency event affecting Changineers infrastructure and systems.

IT is responsible for recovery of the Changineers hosted environment, network devices, and all servers. The team includes personnel responsible for the daily IT operations and maintenance. The team leader is the IT Manager who reports to the COO.
HR & Facilities is responsible for ensuring the physical safety of all Changineers personnel and environmental safety at each Changineers physical location. The team members also include site leads at each Changineers work site. The team leader is the Facilities Manager who reports to the COO.
DevOps is responsible for assuring all applications, web services, platform and their supporting infrastructure in the Cloud. The team is also responsible for testing re-deployments and assessing damage to the environment. The team leader is the Head of Engineering.
Security is responsible for assessing and responding to all cybersecurity related incidents according to Changineers Incident Response policy and procedures. The security team shall assist the above teams in recovery as needed in non-cybersecurity events. The team leader is the Security Officer.

Members of above teams must maintain local copies of the contact information of the BCDR succession team. Additionally, the team leads must maintain a local copy of this policy in the event Internet access is not available during a disaster scenario.

All executive leadership shall be informed of any and all contingency events. Current members of Changineers leadership team include the Security Officer, Privacy Officer.

General Disaster Recovery Procedures¶

Notification and Activation Phase¶

This phase addresses the initial actions taken to detect and assess damage inflicted by a disruption to Changineers. Based on the assessment of the Event, sometimes according to the Changineers Incident Response Policy, the Contingency Plan may be activated by either the COO or Head of Engineering. The Contingency Plan may also be activated by the Security Officer in the event of a cyber disaster.

The notification sequence is listed below:

The first responder is to notify the COO. All known information must be relayed to the COO.
The COO is to contact the Response Teams and inform them of the event. The COO or delegate is responsible to begin assessment procedures.
The COO is to notify team members and direct them to complete the assessment procedures outlined below to determine the extent of damage and estimated recovery time. If damage assessment cannot be performed locally because of unsafe conditions, the COO is to following the steps below.
- Damage Assessment Procedures:
- The COO is to logically assess damage, gain insight into whether the infrastructure is salvageable, and begin to formulate a plan for recovery.
- Alternate Assessment Procedures:
- Upon notification, the COO is to follow the procedures for damage assessment with the Response Teams.
The Changineers Contingency Plan is to be activated if one or more of the following criteria are met:
- Changineers will be unavailable for more than 48 hours.
- On-premise hosting facility or cloud infrastructure service is damaged and will be unavailable for more than 24 hours.
- Other criteria, as appropriate and as defined by Changineers.
If the plan is to be activated, the COO is to notify and inform team members of the details of the event and if relocation is required.
Upon notification from the COO, group leaders and managers are to notify their respective teams. Team members are to be informed of all applicable information and prepared to respond and relocate if necessary.
The COO is to notify the hosting facility partners that a contingency event has been declared and to ship the necessary materials (as determined by damage assessment) to the alternate site.
The COO is to notify remaining personnel and executive leadership on the general status of the incident.
Notification can be message, email, or phone.

Recovery Phase¶

This section provides procedures for recovering Changineers infrastructure and operations at an alternate site, whereas other efforts are directed to repair damage to the original system and capabilities.

Procedures are outlined per team required. Each procedure should be executed in the sequence it is presented to maintain efficient operations.

Recovery Goal: The goal is to rebuild Changineers infrastructure to a production state.

The tasks outlines below are not sequential and some can be run in parallel.

Contact Partners and Customers affected to begin initial communication - DevOps
Assess damage to the environment - DevOps
Create a new production environment using new environment bootstrap automation - DevOps
Ensure secure access to the new environment - Security
Begin code deployment and data replication using pre-established automation - DevOps
Test new environment and applications using pre-written tests - DevOps
Test logging, security, and alerting functionality - DevOps and Security
Assure systems and applications are appropriately patched and up to date - DevOps
Update DNS and other necessary records to point to new environment - DevOps
Update Partners and Customers affected through established channels - DevOps

Reconstitution Phase¶

This section discusses activities necessary for restoring full Changineers operations at the original or new site. The goal is to restore full operations within 24 hours of a disaster or outage. If necessary, when the hosted data center at the original or new site has been restored, Changineers operations at the alternate site may be transitioned back. The goal is to provide a seamless transition of operations from the alternate site to the computer center.

Original or New Site Restoration
- Repeat steps 5-9 in the Recovery Phase at the original or new site / environment.
- Restoration of Original site is unnecessary for cloud environments, except when required for forensic purpose.
Plan Deactivation
- If the Changineers environment is moved back to the original site from the alternative site, all hardware used at the alternate site should be handled and disposed of according to the Changineers Media Disposal Policy.

Testing and Maintenance¶

The COO and/or Head of Engineering shall establish criteria for validation/testing of a Contingency Plan, an annual test schedule, and ensure implementation of the test. This process will also serve as training for personnel involved in the plan’s execution. At a minimum the Contingency Plan shall be tested annually (within 365 days). The types of validation/testing exercises include tabletop and technical testing. Contingency Plans for all application systems must be tested at a minimum using the tabletop testing process. However, if the application system Contingency Plan is included in the technical testing of their respective support systems that technical test will satisfy the annual requirement.

Tabletop Testing¶

Tabletop Testing is conducted with the primary objective of the tabletop test to ensure designated personnel are knowledgeable and capable of performing the notification/activation requirements and procedures as outlined in the CP, in a timely manner. The exercises include, but are not limited to:

Testing to validate the ability to respond to a crisis in a coordinated, timely, and effective manner, by simulating the occurrence of a specific crisis.

Simulation and/or Technical Testing¶

The primary objective of the technical test is to ensure the communication processes and data storage and recovery processes can function at an alternate site to perform the functions and capabilities of the system within the designated requirements. Technical testing shall include, but is not limited to:

Process from backup system at the alternate site;
Restore system using backups; and
Switch compute and storage resources to alternate processing site.

Work Site Recovery¶

Changineers’s software development organization is a distributed team who work from multiple locations with Internet access and do not require an office. Whilst individuals may be affected by disasters, we are resilient through no two individuals being in the same place at the same time.

Application Observability¶

All applications developed by Changineers are fully observable in all environments. By using the latest technologies we can provide active tracing, log aggregation, and real-time metrics for all our systems continuously.

At all times Changineers’s systems have:

Traces sampled using AWS X-Ray which provides deep inspection capabilities for every transaction that flows through the system. Real-time distributed tracing exposes request latency through each layer of the stack, as well as surfaces errors from individual components. Traces are tagged with tenant and user identifiers to allow pin-point support for issues.
Metrics gathered for all important aspects of the system using Amazon CloudWatch. Request latency is captured at p50, p90, and p99 thresholds for all HTTP servers and databases. Success and failure metrics are captured for all API calls. Storage amounts, free space, memory consumption, and invokation durations are all monitored.
Logs generated for all transactions and stored in Amazon CloudWatch Logs, which contain detailed information about the activity of the system. No sensitive data is ever stored in log files.

The Changineers Development team has access to various Amazon CloudWatch Dashboards that provide visilibity into the health of the system at all times.

Alerting¶

When any of the aforementioned traces, metrics, or logs indicates that the system may be performing in a suboptimal manner (errors, performance degrading, suspicious or malicious activity) an CloudWatch Alarm will trigger and notify any subscribers (see below).

Examples of Alarms:

A Lambda function invokes, but fails before completing
A DynamoDB query takes longer to complete than an acceptible threshold
Our CloudFront distribution receives an unexpected spike in activity

Depending on the severity of the Alarm, different actions are taken. High severity issues result in the Head of Engineering and the Development team being contacted immediately via an interruptive medium (e.g. phone call), whilst lower severity issues may trigger e-mails or other non-interuptive methods. The severity of an issue varies based on it’s impact to customers.

Outages¶

Due to the architecture that Changineers Platform uses, it is very unlikely for there to be any large-scale outages. The use of Amazon API Gateway and AWS Lambda with zero-downtime deployments results in few opportunities for service availability to be affected.

In the case of any outages, the Changineers Development team will be immediately notified via our CloudWatch Alarms, and affected customers will be notified.

Production Environments and Data Recovery¶

Production data broadly takes two forms in the Changineers Platform:

User uploaded files such as pictures, videos, and documentation they produce.
System generated data through the use of the Platform, such as user activity data, textual responses to questions, timestamped video annotations.

User uploaded files¶

For user uploaded files, data is stored in Amazon S3 in the tenant’s preferred region. This type of data is never deleted by the application (prevented by IAM policies), and can only be deleted by priviliged administrator users with the explicit requirement of multi-factor deletes. All modifications to files are versioned, and over time files are archived to Infrequent Access storage, and eventually to Amazon S3 Glacier for long-term storage.

Data in S3 buckets are distributed across multiple data centers within a single region. Any failure of a single datacenter does not affect the availility of data.

In the event that archived data is required to be retrieved, a request to Glacier can be submitted.

System generated data¶

For any data that is created by the system on the behalf of a user, for example any data typically stored in a database, the Changineers Platform stores the data in Amazon DynamoDB. Data in DynamoDB is automatically replicated across multiple data centers within a tenant’s region, providing resilience against data center failures.

Additionally, our DynamoDB is configured to take continuous snapshots of the data providing to-the-second backups. Periodic backups are also taken, and can be backed up to secondary regions if the tenant’s data sovereignty requirements permit.

Recovery¶

Recovery of production Environments and data should follow the procedures listed above and in Data Management - Backup and Recovery.