Ansible/Docker/Kubernetes Environment (ADK) – Simplifying cluster environments, minimizing maintenance, and enabling smooth upgrades is critical to analytics and IoT clusters. Finding professionals to manage these clusters is difficult and keeping these engineers trained on the latest updates is an issue. This is particularly hard when running multiple applications on a cluster which may interact and require specific security implementations. Paladian has developed the ADK environment to make deploying and managing these platforms as simple as specifying the requirements in a unified spreadsheet format and linking to the custom configuration features of choice. Updating the environments is as simple as modifying parameters in the configuration files and running an update ADK script, to orchestrate a simultaneous update for all the relevant applications in the preferred order.
The Paladian SureConnect™ software ensures your data is collected securely by tracking the robustness of the cluster fabric and the health of its Ethernet and PCIe connections. During routine maintenance in a data center, cabling can be damaged. SureConnect™ periodically checks that the connections are operating properly within the transmission guidelines. If SureConnect™ determines a link has errors or a high rate of re-transmissions it will automatically transfer the connection to a backup path and notify the administrator. In large clusters the application software may cause a server to go down or disconnect. The cluster servers typically manage their own data. It can be time consuming to reconnect to the data or recreate stale data once a server is restarted. In case of large drives this can take several days before the system is stable again. SureConnect™ allows this storage to be reassigned automatically to another hot server so that there is no noticeable effect on the cluster’s operation and risk of additional errors causing a total application shutdown.
The Paladian Watchman™ software monitors your IoT data flows to ensure normal measurements or events and that your sources continue to operate within requested performance parameters. This module uses machine learning routines to establish baselines, tracks that periodic measurements or events are returned regularly, verifies that the data is within bounds using the established baselines, highlights any anomalies, and determines if that the data is trending towards an excursion – results that often are difficult to observe when running routine dashboards or reports. The Paladian machine learning routines (Watchman™) provide streaming analysis, data consolidation/deduplication and monitoring.
Many companies keep up to 90 days of hot/warm data on-premises due to response times or bandwidth concerns. However, to minimize long-term support infrastructure, longer-term data may be moved to either a local or external cloud object store, like Amazon’s S3. As a compromise some parts of longer-term data can also be kept locally by using a local cache and just keeping additional copies in the cloud for reliability. The HSM (Hybrid Storage Manager) enables customers to easily set up a hybrid storage environment using custom specific requirements. Once HSM establishes an environment, customers can access their data as needed, monitor performance and cost parameters, and fine tune the settings to optimize the performance to the use models.
The PD2001-SPL is a complete Splunk cluster built up with a high performance NVMe storage platform (PD2001), 1u indexers, search heads, and supporting services. It includes a Paladian Cluster Manager (PCM) along with a single pane of glass health monitor, and Redfish and NVMeoF support. The Splunk services are deployed using Ansible playbooks defined with an easy-to-use Splunk Deployment Tool. Each server is set up with Docker containers and is managed by a Kubernetes resource manager ensuring seamless future growth. One PD2001 cluster storage unit can offer 33M IOPS and 720TB in 2u – allowing 10x query performance in 20% of the standard rack space. An optional 1u cold storage unit provides increments up to 960TB per 1u.
Paladian offers single button deployments, out of the box operation, industry leading performance and form factor, and simplified health management. The PCM provides 24-hour health monitoring and alerting. Add-on modules offer automatic server and connection fail-over, encryption key management, and hybrid storage support.
The PD500 and PD2001-IOT appliances provide powerful edge platforms for both mobile environments and small data centers or wiring closets. The mobile PD500 provides up to 10M IOPS in a half rack width unit with up to 360TB. Both include IoT SureCollect™ software to ensure your IOT data is collected securely and your sources continue to operate within requested performance parameters when serving as an IoT sensor. The Paladian machine learning routines (Watchman™) provide streaming analysis, data consolidation/dedup and monitoring, available in the end of 21'Q2.
For next generation IoT networks where IoT sensors are embedded in edge switching/routing equipment as with Sentryo (Cisco) and Claroty, the Paladian platforms act as upstream data collection points replacing last generation data historians and data middle ware tools. Paladian provides longer data retention and tight integration with SIEM tools in the SOC simplifying investigations or remediation.
Our storage architecture is optimized to fully utilize the next generation of storage (NVMe) so your applications can deliver on all their promises. Our NVMe flash array delivers 33M IOPS and 120 GB/s in two rack units, supporting NVMeoF to enable customers to use standardized management tools. The architecture is 100% linearly scalable to PBs of storage without any compromising storage controllers, nor additional protocol latency.
For mobile edge applications Paladian offers the PD500 which provides a quad processor platform integrated with 12 NVMe SSDs which can be connected to up to 40 external laptops or direct IoT devices collecting remote data. The data from these external units can be processed and then stored on the NVMe storage for future transfer or transmission. Look for our 5G cellular add-on card next year. This is ideal for IoT consolidation points, autonomous vehicles, or real-time drone video capture & analysis.
For applications like Splunk we can deliver solutions with 10x-100x more performance while reducing the TCO over 50%.