EUVIMEDEUVIMED entdecken
Crossref

From Application Server to Pod: A Readiness-Scoring Framework for Migrating Legacy J2EE Middleware Workloads to Enterprise Kubernetes Platforms

Kartheek Reddy Jinna, Sundara Vadivel Thilagaraj

International Journal of Engineering & Extended Technologies Research · 2022

DOI: 10.15662/ijeetr.2022.0403006 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Engineering & Extended Technologies Research, 2022. Verfasst von Kartheek Reddy Jinna und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Regulated enterprises continue to run revenue-critical services on application-server topologies that were assembled over two decades: clustered WebSphere Application Server Network Deployment cells, WebLogic domains, and IBM MQ queue managers bound to those servers through JNDI, XA transactions, and node-local configuration. As these organisations adopt enterprise Kubernetes platforms, the choice of how each application should move is frequently made case by case, and the resulting migrations stall, overrun, or are rolled back after cutover. This paper proposes a readiness-scoring framework that makes the decision explicit and repeatable. Each application is scored on six dimensions (session state, JNDI/JDBC coupling, messaging dependency, clustering model, configuration externalization, and startup behaviour) using an evidence-based rubric. A weighted score on a 0 to 100 scale, combined with three hard gates, places the application in one of three bands that map to a migration pattern: rehost in a container on the same runtime, replatform to a lightweight runtime such as Open Liberty or Tomcat, or decompose incrementally behind a strangler facade. We derive the framework from practitioner experience with clustered middleware estates in financial services and evaluate it on a representative portfolio of 40 applications. Migrations whose pattern agreed with the framework had a first-cutover rollback rate of 6.3%, against 50% for the eight decisions that disagreed, and roughly half the escaped defects per application. Band assignments were stable under weight perturbation, which suggests the rubric, not fine weight tuning, carries most of the signal

Zitieren

APA

Jinna, K. R., & Thilagaraj, S. V. (2022). From Application Server to Pod: A Readiness-Scoring Framework for Migrating Legacy J2EE Middleware Workloads to Enterprise Kubernetes Platforms. International Journal of Engineering & Extended Technologies Research. https://doi.org/10.15662/ijeetr.2022.0403006

Vancouver

Jinna KR, Thilagaraj SV. From Application Server to Pod: A Readiness-Scoring Framework for Migrating Legacy J2EE Middleware Workloads to Enterprise Kubernetes Platforms. International Journal of Engineering & Extended Technologies Research. 2022. doi:10.15662/ijeetr.2022.0403006

BibTeX
@article{Jinna2022,
  author = {Jinna, Kartheek Reddy and Thilagaraj, Sundara Vadivel},
  title = {From Application Server to Pod: A Readiness-Scoring Framework for Migrating Legacy J2EE Middleware Workloads to Enterprise Kubernetes Platforms},
  journal = {International Journal of Engineering & Extended Technologies Research},
  year = {2022},
  doi = {10.15662/ijeetr.2022.0403006}
}
Kostenlos starten →