9. Pro Tip
HDFS Vs.
Hadoop Vs.
Application
Loader
Big Bulks
Bring your
files to your
cluster
Load from
Several Nodes
Convertro
Pro Tip Pro Tip
Vertica The Convertro wa
33. Convertro
Many
Deletes / Updates
Node
Crash
Slow Recover
Process
Checking Recovery
Status
Incremental
recovery
replay-
delete
Solution 3
Delete only
one file
Incremental By Containers
Vertica The Convertro wa
The data for the model is in vertica
Process in R
Attribution run on hadoop based on the model that R calculated .
Vertica :
Raw : 150TB
DB Size : 60 TB
Top Clients: (top 8 above 1 TB )
Intuit 3 TB
qvc 2 TB
Dashboard :
qvc => 260 GB => 8 Billion rows
intuit => 200 GB => less than 8 Billion rows Vertica scan it and fetches a result in less than a second
Complex analytics on a year takes about 5 seconds.
Facebook 35 TB per hour 2 years ago !!
150K rows per sec
Regular days 10 – 15 billion rows per day
40 billion rows per day
New vertica feature => Vhash
out of the box improvements =>
denormalize => data model changes are Game changer !!!.Vertica can handle big joins => merge joins
out of the box improvements =>
denormalize => data model changes are Game changer !!!.Vertica can handle big joins => merge joins
MMM => measure measure measure => data collector tables .
out of the box improvements =>
denormalize => data model changes are Game changer !!!.Vertica can handle big joins => merge joins
MMM => measure measure measure => data collector tables .
out of the box improvements =>
denormalize => data model changes are Game changer !!!.Vertica can handle big joins => merge joins
MMM => measure measure measure => data collector tables .
out of the box improvements =>
denormalize => data model changes are Game changer !!!.Vertica can handle big joins => merge joins
MMM => measure measure measure => data collector tables .
out of the box improvements =>
denormalize => data model changes are Game changer !!!.Vertica can handle big joins => merge joins
MMM => measure measure measure => data collector tables .
great database => out of the box performance
even lebron => don’t put billing on it
right tool => hadoop loader,extended analytics ,flex table,udx
Keep it simple => easy to debug easy to maintain
great database => out of the box performance
even lebron => don’t put billing on it
right tool => hadoop loader,extended analytics ,flex table,udx
Keep it simple => easy to debug easy to maintain
great database => out of the box performance
even lebron => don’t put billing on it
right tool => hadoop loader,extended analytics ,flex table,udx
Keep it simple => easy to debug easy to maintain