In architecting the POSSE Cloud environments, we recognized that systems may fail due to various factors, and clients must be able to continue working when this happens.
To deal with this reality, POSSE Cloud environments are designed with data and system services redundancy. This is further supplemented by running client systems on highly available Kubernetes clusters where workloads are dynamically moved within the cluster without impacting live client activities. Further, each cluster is built with enough excess capacity (compute, memory, and storage) so that any host node can be down, and the remaining host nodes have sufficient reserve to continue operating without client impact.
The Oracle database is at the heart of the POSSE Cloud solution and contains all structured data and configuration metadata. Backing up the POSSE database secures all the information and metadata stored within it. All completed transactions on the database (inserting data, changing data, deleting data) are logged to the Oracle transaction log (Oracle Redo Log). Three independent copies of the logs are retained on separate disks to reduce the risk of data loss. Data stored as part of POSSE Cloud agreements belongs to the client, and its safekeeping is of the utmost importance to Computronix. As such, Computronix has architected a backup solution that meets industry best practices using Microsoft Azure.
Azure Site Recovery (ASR) is configured to
continuously replicate the Oracle services
(including data) to the Azure Disaster
Recovery (DR) site. The continuous nature of
Computronix’s ASR configuration ensures that
the transactions are streamed to the DR site
as they are written to the Oracle transaction
log files. This ensures that data is available
on the DR site within the Recovery Point
Objective (RPO) of 4 hours.
Using ASR configuration Computronix has
found that we are able to exceed the RPO and
transactions are frequently available in DR
within minutes of the completion of Oracle
processing.
The health of the replication process
is monitored and configured to alert if
it becomes unhealthy. Remediation of
unhealthy conditions is initiated promptly.
At least daily, Oracle Recovery Manager
(RMAN), the industry-standard utility for
backing up Oracle, autonomously creates
a backup on Geographically Redundant
Storage (GRS). GRS provides replication
between the primary data center and the
DR data center and delivers reliability of
16 9s (99.9
POSSE Cloud recognizes that there are many aspects to securely protecting our customers business. This includes Role Base Access, Privileged Access Management, remediating vulnerabilities, and many more.
To address these and many other security concerns, POSSE Cloud is designed, implemented and maintained to be compliant with NIST 800-53 Moderate assessment using FedRAMP Moderate parameters. To ensure compliance with this standard and to be transparent with our customers, POSSE Cloud engages a FedRAMP certified auditor (3PAO) to complete an annual audit.
PADD
The POSSE Alternate Document Datastore (PADD) is a repository for documents. Similar to Oracle data storage, PADD uses GRS storage, to achieve 99.99999999999999% file reliability over a
given year.
Encryption
Data at rest and data in transit is encrypted using FIPS 140-2 algorithms. This ensures that sensitive
data is not available to unauthorized entities. It also ensures that data has not been altered while in transit or at rest.
Ensuring continuity in client operations in the event of loss of the primary hosting site is not only a key provision within POSSE Cloud agreements, it is also a primary objective of the Hosting Team. To that end, each client is provided with a geographically distinct disaster recovery environment in a Microsoft Azure data center completely independent of the primary hosting site.
In compliance with NIST 800-53, Computronix performs annual testing of the DR process including testing of client data recovery and system functionality.
