Architecture
Field notes, architecture deep-dives, and practical guidance on enterprise storage, data protection, and infrastructure - written by the engineers who build it.

Retiring an old storage array gets a lot harder when the servers involved boot straight from SAN. Here's how Hitachi's Universal Volume Manager lets you virtualize the old array behind the new one, cut hosts over in minutes, and migrate the data quietly in the background.
.png)
Every enterprise has that one server nobody will touch — end-of-life OS, vanished vendor, and a reboot history scary enough to keep it out of every patch cycle. This post looks at why these systems accumulate, why "leave it alone" is an illusion that AI-accelerated attacks are dismantling fast, and what actually works: airlocked networks and load balancers to buy time, third-party support to extend the runway, and ultimately a migration approach built on clones, isolated testing, and rollback plans — so getting off the legacy box becomes routine instead of a sev 1 in waiting.