Explain like I'm five · Cloud

What is Edge and Fog Computing?

Edge and fog computing, told as a story about a local branch — and a hub in between — handling things before the far-off HQ has to.

HQ building (far away)a long walk to get therea visitor from across town takes forever to reach HQ

A visitor from across town takes forever to reach HQ.The farther HQ (the central server) is, the longer a visitor (a request) waits.

everyone funnels into the same far-off HQfunnel everyone into one HQ, and the line adds to the long trip

When every visitor funnels into one HQ, everyone waits longer.Visitors' requests need to be handled somewhere closer.

HQfog: middle hubedge: local branchedge = the local branch, fog = the hub between it and HQthe branch handles requests near the visitor first, so the reply comes faster than from HQ

Edge = the local branch, fog = the hub between it and HQ.The branch handles requests near the visitor first, so the reply comes faster than from HQ.

The branch nearest the visitor handles the request firstedge = the local branch
Fog bundles several branches togetherit's the hub between branch and HQ
The closer the processing, the faster the responseresponse latency drops
Branch data still has to sync with HQ eventuallythat's called synchronization
where requests get handled firstedge (branch): simple requests, handled right therefog (middle hub): gathers and sorts data from several branchesHQ: handles full storage and heavy analysisthe closer the processing, the faster the response

Edge, fog, and HQ each take on different work.Nearby branches take simple requests; HQ takes heavy analysis.

the branch is too small to hold all the databranch data still has to sync with HQ eventually

The branch is too small to hold all the data.Branch data still has to sync back up with HQ eventually.

In one breath

Edge computing means handling things right at the nearby local branch. Fog computing is the hub in between that bundles those branches together. Both spare requests the full trip to the central HQ, shortening the distance traveled.

Edge computing processes data right next to where it's generated (e.g. right beside an IoT sensor). Fog computing adds a layer between the edge and the central cloud, bundling multiple edge nodes together. Both cut latency and reduce load on the central network, but HQ — the central cloud — still has far greater processing capacity and storage reach.

When grown-ups say it

Edge computing
Processing right at the small local branch. Right next to where the data is generated.
Fog computing
The middle hub bundling several branches. One more layer, between edge and HQ.
Latency
The wait that grows with distance. Shorter, the closer the processing happens.
Synchronization
Matching branch data back up with HQ. Needed because the branch alone can't hold it all.
IoT
The devices out in the field generating data. Edge computing processes these devices' data right beside them.
Central cloud (HQ)
The HQ building handling full storage and heavy analysis. The very building this whole book is set in. → the shared office building
Serverless
The meeting room that opens only when used. Serverless functions are sometimes placed at the edge too. → the hourly meeting room
Auto scale
Automatically adding and removing servers. The server count inside an edge branch can also scale with demand. → bigger office or more offices
Next story
The Line That Splits Two Groups the Widest →
Want another term as a picture book? Request →