Jelastic Adds High Availability to its Java PaaS, Cloud Hosting

Jelastic, creators of a PaaS (platform-as-a-service) infrastructure for Java application development has added new features to ensure automated high availability of Java apps. Jelastic lets devs run any Java app in the cloud without code changes and or rewrites for APIs, and offers a full range of support for creating, deploying, scaling and managing apps.

Tags: paas, platform as a service, Java, high availability, clustering, SDLC,

jelasticcloud_02Jelastic, creators of a PaaS (platform-as-a-service) infrastructure for Java application development has added new features to ensure automated high availability of Java apps.

 

Jelastic was built to help devs focus on innovation and writing code, and avoid the need to spend time and energy managing and configuring servers, or worry about infrastructure problem, according to Ruslan Synytsky, founder and CEO of Jelastic.

 

Jelastic runs any Java app in the cloud without code changes and without writing for specific APIs, and already supports a full range of simple to use features for creating, deploying, scaling and managing applications, according to Synytsky. In specific, devs using Jelastic have a wide selection stack components, including application servers (Tomcat, GlassFish and Jetty), databases (MySQL, MariaDB, PostgreSQL, MongoDB), nginx load balancer, and Maven build system,. This wide choice of components, in turn, helps ensure applications will run and scale easily on Jelastic platform, he added. 

 

Jelastic’s latest HA (high availability) features are simple to access.  When devs click a button, from within the environment configuration, Jelastic will then clusters each application server with another application server running on a separate virtual machine and replicates sessions between them, Synytsky said. 

 

Jelastic also will automatically configure the load balancer, including the enabling of cookie-based/sticky sessions. By this approach, if one application server fails or is taken offline for maintenance or application upgrade, the other machine keeps running off the stored sessions it was sharing within the cluster.

 

An excerpt from the Jelastic blog describes the process and the technology is more detail, with an example of how it provides high availability for GlassFish.

Up until now, GlassFish could be used as a separate server within Jelastic, but now you have the option for high-availability for GlassFish, allowing for full instance and cluster replication.

 

We kept the native GlassFish clustering architecture, based on the concept of and administrative domain. Administrative domains consist of clusters and instances, which are controlled by DAS (Domain Administration Server).

 

 You can manage the central repository using the Admin Console. It is an easy to use GUI which supports all features available in GlassFish.  DAS manages Java instances in the domain. GMS (Group Management Service) is responsible for providing the information about a cluster and its instances. 

 

Jelastic also takes care of session replication in GlassFish

Instances in clusters are paired up. If one of the instances fails, the users who were on that instance get automatically switched to the other instance in the cluster. The replicated instance has all of the sessions that were in the failed instance: end-users never notice any change. If for some reason both instances in a cluster fail, users get redirected to another cluster. Jelastic uses NGINX as a load balancer, which can redirect users’ requests when instances fail.

All these services come without vendor lock-in, he said. “Jelastic’s technology makes all of this possible and it is as easy as pointing and clicking,” Synytsky added.

 

The automated vertical scalability feature within Jelastic adds the memory and processing resources to the machines when necessary so if one server goes down, the alternate server not only gets the sessions of the failed server, but also the CPU and RAM required to serve them, officials said. Because the process happens automatically, the end-users are not even aware that anything has happened. It also mitigates human error and enables developers to focus on creating apps.


back

Share
Go