Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
SlideShare a Scribd company logo
for SEC 2016
.io
i1
onthe.io
Den Golotyuk Big data from 30 million daily users
Our figures
1 year from zero
30m+ uniques across our customers apps
200 customers from USA/Europe/Asia
Big Data
i2
Huge data flow
Den Golotyuk Big data from 30 million daily users
i4
No apps anymore
i5
One source a year
Fantastic AI possibilities
Issues
Too complex tools...
Tons of software addressing data collection and
processing. Universal. Hard to setup. Hard to use.
Den Golotyuk Big data from 30 million daily users
Den Golotyuk Big data from 30 million daily users
It show pageviews
Everything works already
i10
The gap
Analytics
i11
Bias
i12
Entropy
So we just get data
do the complex stuff inside
and get the simple answers outside
Den Golotyuk Big data from 30 million daily users
Behind us
2b daily request, 100Gb daily gzipped backups, 150
cloud and physical nodes
4 engineers
Microiterations. Microservices. Metrics driven.
i13
Stack
Node, JS, PHP, Python, Vertica, Mysql, Mongo,
Sphinx, Gearman, Redis, Statsd
Cloud
What we learned
1. We’re dumb
Hadoop, Spark, Cassandra, Elastic...
too advanced for us
2. Microservices
as a management approach
3. Expensive systems
aren’t cool, they’re poor
4. Most businesses
don’t even know what questions to ask
5. Nothing is really done
in analytics, so join us
den@onthe.io
.io

More Related Content

Den Golotyuk Big data from 30 million daily users