Google particulars three frequent mainframe modernization errors


As extra organizations embark on mainframe modernization journeys, Google needs to ensure they head down the best path. The corporate outlined frequent pitfalls and antipatterns companies face when migrating or modernizing their workloads. 

“Migrating or modernizing your mainframe workloads is advanced and difficult, even below ideally suited circumstances,” Travis Webb, options architect at Googe, wrote in a weblog submit. “Should you keep away from the antipatterns mentioned on this doc, you enhance the percentages of a profitable transformation.”

RELATED CONTENT: Mainframe for DevOps places an finish to silos

Whereas these approaches may match for some circumstances, Webb warns in opposition to them as a result of “they’ve a excessive likelihood of failure.”

In accordance with Webb, the three commonest antipatterns are: 


Large bang rewrite purposes: Rewriting or re-architecting your legacy mainframe code right into a extra trendy language or design patterns will help pace of utility improvement and future-proof options, however it’s a capital-intensive and time-consuming endeavor, Webb defined. Dangers embrace price range overruns, unanticipated complexity and workers turnover. “Even for corporations which have the tenacity to see by means of a multi-year transformation effort, the uncooked value of a rewrite is usually prohibitive. When in comparison with all different approaches, a giant bang rewrite is the most expensive approach to modernize your mainframe software program,” Webb wrote. 

Raise-and-shift migration antipatterns: Organizations are sometimes tempted to maneuver an utility from one system to a different, as an illustration transferring the mainframe into the cloud. This generally is a fast approach to get away from an on-premise setting, however organizations nonetheless stay locked into the mainframe ecosystem and depending on an emulation layer. “That dependency may end up in a brand new set of technical challenges. Challenges which are typically unfamiliar to the groups sustaining the mainframe software program. Unfamiliarity can result in further reliance on a brand new, single-vendor cloud ecosystem,” in accordance with Webb. 

In-place modernization antipatterns: With this antipattern, as a substitute of rewriting and re-architecting mainframe code, you deal with the standard, maintainability and testability of software program — however you continue to are subjected to the identical dangers of a giant bang strategy. “Any strategy involving manually updating your mainframe software program can have price range and time constraints. These efforts additionally typically endure from the second-system impact. Efficiency and correctness points inevitably come up as a result of rewriting enterprise logic in a brand new language requires in depth testing earlier than it aligns with the earlier performance,” Webb wrote.