Use bare metal server hosting to separate Keycloak runtime, proxy, and database capacity, then test login bursts, session growth, upgrades, and recovery against attributable resources.
Reserve CPU and memory for login bursts, token operations, JVM pressure, and active sessions under representative tests.
Match SSD or NVMe capacity to identity data, transaction pressure, backups, and restore operations under load.
Use separate nodes when instance failure or maintenance must not interrupt login across connected applications.
Your team owns Keycloak, proxy, database, upgrades, monitoring, backups, restores, and incident response procedures.
Centralising SSO in Keycloak concentrates login bursts, token issuance, active sessions, federation lookups, and administration on one production dependency during peaks and maintenance windows. Capacity planning must cover the runtime, reverse proxy, production database, backup work, and recovery objectives—not just average request volume.
At Melbicom, we provide isolated CPU and memory, SSD or NVMe storage, bandwidth choices, and regional placement through ready-to-deploy or custom dedicated infrastructure. Teams can test authentication demand against attributable resources, then assign separate nodes when instance failure or maintenance requirements justify a clustered design.
Your team keeps the software boundary: realms, clients, identity providers, authentication flows, TLS, database topology, monitoring, upgrades, backups, restores, and failover procedures. We provide 24/7 technical support for the infrastructure; your team distinguishes resource contention from software faults and retains incident ownership during recovery.
Size runtime and database capacity for employee login bursts, active sessions, federation lookups, and administration.
Keep token issuance, session activity, and identity data attributable as customer demand changes across applications.
Reserve CPU, memory, and database I/O for directory lookups and external identity provider traffic under load.
Separate client, realm, and authentication-flow changes from infrastructure capacity planning and related recovery tasks.
Run a single instance when continuity requirements permit, with health checks, backups, and tested recovery steps.
Run Keycloak on separate nodes so login can continue through instance failure, maintenance, and controlled upgrades.
Size storage and database I/O around backup volume, restore work, latency targets, and recovery objectives.
Run upgrades through health checks, rollback plans, configuration control, and clearly assigned incident ownership.
