Why #Webscale Reason #2: System Design #Nutanix

Keeping things up to date in the Enterprise usually happens when a support contract is about to run out or when a new feature is needed. Teams are reluctant to make changes when outages need to be planned, load on the systems needs to low and possible manual intervention is needed throughput the process.

Web-Scale allows non-disruptive rolling upgrades, no single point of failure and minimal intervention. Self-defining (and versioned) disk format, like protobufs This way, upgrades of disk data can be done lazily. Web-scale can’t assume 1-shot data upgrades, given the scale that it could consume. This also inherently allows client/servers relationship to be a different versions throughout the process which is a pretty big accomplishment onto itself.

Join Web-Scale Wednesday to learn more and hear from Industry leaders like Dmitriy Ryaboy Engineering Lead at Twitter.


Speak Your Mind