Tim Berglund
Commentaire de texte : Tim Berglund. Recherche parmi 300 000+ dissertationsPar neabog • 18 Avril 2015 • Commentaire de texte • 903 Mots (4 Pages) • 732 Vues
This presentation is done by Tim Berglund for Oreilley’s network interviews.
Tim is a teacher, author, and technology leader with DataStax, where he serves as the Director of Training.
He explaining how distributed storage doesn’t have to be a confusing issues:
Is simple as a coffee shop.
But first he explains few different types of distributed storages strategies:
Read replication
Sharding
Consistent hashing
Distributed filesystems (HDFS)
and explained what is existing right now like MangoDB, Cassandra…
He took the example of the coffee shop to make it as clear as possible:
He started first with a brief description of a non-distributed storage system, according to coffee shop terms used.
In his words: « You walk into the coffee shop, introduce yourself to the barista, who takes your name and order and dutifully writes it down. Now every time you enter the coffee shop, the barista recognizes you, looks down at her paper and knows exactly what you want. »
This is considered as the best storage method it says because "it's possible to pack all kinds of sophistication into that little drum and there are all kinds of things you can have the database do when it's in control of the whole world."
In this « model » the environment is still the same: you and the barista.
According to evolution of dates this is going to be inefficient as data for the future he said.
Berglund introduces over here the usage of the distributed storage.
The read replication: the barista is the main customer/order list chief. He provides a copy of his list to the others barisata’s « the helpers » who can do the same job as main barista because of the knowledge they have of what you would like if you jump in to the coffee shop. Data by this way is replicated to the other helpers. This can solve problems, Berglun said, and make our business grow.
At the same time this replication of data can insert new problems.
He takes the example when you can change your main order, so the main barista will update your list, and he will have to make new copies for each of the helpers. This can introduce some time delays.
"This is what we call an 'eventually consistent' system, »
"It will catch up … but I have some period of time that's going to elapse before I can see that."
Considering Berglund’s terms:
« you just have to know it’s happening » he said. He’s pointing out over here the benefits of read replication.
If the coffee shop continues to rise the single barista is not going to head up everything so he has to divide the master information’s he’s got to the other baristas.
...