Last quarter we walked into a “brand-new” data-centre build that had stalled for weeks. The root cause – wrong switches, wrong topology.
▪️ Juniper Networks QFX series (already end-of-life) were picked for leaf duty – great price per port, but no VXLAN routing.
▪️ To make matters worse, the team linked the two leafs together and gave each leaf only a single uplink to the spine. Result: uneven traffic, unpredicatble resilience and higher long-term costs.
𝐅𝐢𝐱:
After a fast HLD/LLD workshop we swapped to Arista Networks leafs running Broadcom Trident2+ ASICs and rebuilt a true spine–leaf fabric with dual-attached leafs and EVPN-VXLAN multihoming. Racking to production in days, not weeks, no more head-scratching.
𝐅𝐚𝐦𝐢𝐥𝐢𝐚𝐫 𝐬𝐭𝐨𝐫𝐲 #𝟐
An ISP client bought a pair of Juniper Networks MXs two years ago – enough ports then, but no scale headroom. Subscriber growth + new peers = hardware now on life-support and headed for storage. The refresh bill? Way higher than planning the right chassis from day zero.
𝐓𝐚𝐤𝐞-𝐚𝐰𝐚𝐲
Cheap today can be expensive tomorrow. Architect with scale and silicon in mind, backed by a solid HLD/LLD, and the implementation just… works. Spend a week on documentation and save months of re-cabling and CapEx.
Have you already battled a deployment that dragged on for weeks and don’t want to step on the same rake again? Or are you kicking off a new build and want to avoid that pain altogether?
Get in touch with our Sales team or drop us a line at [email protected] – we’ll help you nail the design and rollout (and can implement it for you, too).
