Month: October 2011

electron delivery vehicals  2500 data points  per vechical,   real time data collection when they are on

US dept. gave 32 million dollar to collect the data to learn about energy use

uses  stormMQ  as queuing system.

realtime use batter use  charge, location

rapid large batch very quickly

DBA  human working on mysql,  each truck has own data base  and cluster of dbs in cloud

call data to UI  10 20 seconds a page,  used caching to improve

If too much data coming in, or has not change they will not insert to save.  But then has to write code to fill the gaps with values

good talk

History of REST

Constraints are a good thing

REST is an architecture  not a plugin

goals

1.  scale

2. general UI

3. Independent component

4.

Constraints

Client save server

Stateless (never store anything server side)

Caching

Layersystem  (allow to put in intermediaries eg. load balancing, caching  is to help scale.)

Code on demand  ()

Uniform interface (4 guiding principals  a, uri, )

URL to access e.g media type get back url to get one pint on api to the next)

http   browser identifies what comes back.  could be html2 to 5

Versioning:  web does not matter on the web  but make sure the links are still valid.