piXserve-Ecl and piXserve-Hcl for Hadoop Clusters



piXlogic has partnered with Cloudera and Intel in order to deliver best-of-breed image and video analytics at scale to big-data customers.  Through the standard Cloudera Manager interface, piXserve can now be deployed "on-the-fly" to worker nodes within an hadoop cluster.  piXserve-Ecl and piXserve-Hcl are special versions of piXserve that work with image/video files stored in HDFS (Hadoop File System), and interface with the Cloudera software platform so that the application can be deployed, managed, and monitored through the standard Cloudera Manager interface.




This provides significant scalability under control by the customer's system administrator, through the standard Cloudera Manager interface.  Running piXserve on the cluster is as simple as deploying the piXserve parcel on the worker nodes you have defined.

Read the Cloudera white paper:  Working with Images and Videos at Scale.


Architecture Summary

A typical deployment in a Cloudera cluster involves multiple piXserve-Hcl instances deployed on worker nodes within the cluster, and one or more piXserve-Ecl manager instances installed either within or outside the cluster.  piXserve runs as a service managed via the Cloudera Manager (CM) interface.  The software is deployed as a "parcel" via the CM.  The administrator defines how many piXserve worker nodes to deploy in the cluster, and all functions including monitoring and security are managed in the standard way via the CM.   a clustervia the The manager node queries name-node(s) through Clouder


Comprehensive Search Environment

Users running piXserve on Cloudera clusters have access to the standard web-browser interface (desktop and mobile).  Navigating and searching for faces, text strings, objects, scenes, etc. in your big-data environment is now easy and straight forward.  Furthre, through the piXserve Web-Services API (REST based) you can put your images and videos to work by discovering patterns and correlations with other data types/sources in your cluster.  Depending on your mix of data and use case scenario, this could expedite the work of your analysts, further improving their productivity and effectiveness.  

Please contact piXlogic for additional information.