13
HuMongous DB Back to Documents @Name => Vivek Parihar @Github => vparihar01 @Twitter => vparihar

Hu mongous db v2

Embed Size (px)

DESCRIPTION

This presentation is on MongoDb basic , it is short and covered a lot of enlightened point which is really helpful for the people who are novice in MongoBb. Here we not only discuss the importance of MongoBb, but also a different scenario where SQL and No SQL (i.e No Only SQL) is good and each one possess their own importance. Also some info about 10Gen's support for MongoBb users.

Citation preview

Page 1: Hu mongous db v2

HuMongous DBBack to Documents

@Name => Vivek Parihar@Github => vparihar01@Twitter => vparihar

Page 2: Hu mongous db v2

What's it About

No Strings AttachedNo Relations , No Craps,No Breakups (I meant System Crashes)

Page 3: Hu mongous db v2

But we do believein Pairs....Key and Value only!

What's it About

Page 4: Hu mongous db v2

Scale! Grows faster than Indian population!

What's it About

Page 5: Hu mongous db v2

Say goodbye to speed bumps like with MySQL, those painful virtual tables and clutter of indexes!

MySQL.... Naaaah

Page 6: Hu mongous db v2

Low retrieval complexity. Woah! No more terrible joins! All hash key pairs doing the trick. And did we hear you say caching! MongoDB backend …. JUST DOESN’T REQUIRE IT.

MySQL.... Naaaah

Page 7: Hu mongous db v2

Object Modeling (“My zillionth ORM layer that only I can understand and maintain”)

MySQL.... Naaaah

Page 8: Hu mongous db v2

We couldn't have had more reasons to adopt it!

Adoption

Page 9: Hu mongous db v2

-Proj 1: We used it for bidding platform- Problem Solved with mongoDB

1) Fast and easy Replication2) No need for joins, transactions .3) Average response time for sending data to server reduced to 60%4) Average response time for displaying data reduced to 50%

Our Experience

Page 10: Hu mongous db v2

-Proj 2: We used it for Logging System(Rinse, repeat :-) )Problems faced with relational dB

1) Logging were slow, stalled or down. 2) Risk of filling up a disk with excessive log information.3) Need to create an index on timestamp.4) Schema changes(Dude don't think of it).

Our Experience

Page 11: Hu mongous db v2

- Breaking relational architecture mindset- Never thought of moving Horizontal with no limit.- to imagine world without joins.-

But it was not easy

Page 12: Hu mongous db v2

-MongoDB is fast no matter how you slice it-Very rapid development, open source-Document model is simple but powerful-Try 10gen guys use cases

Just starting?

Page 13: Hu mongous db v2

Lets take this up on Beer!

Advanced Users