Introduction to Riak

Introduction to Riak
Christopher Meiklejohn
BOBkonf 2015
@cmeik

View Slide

History

View Slide

Published SOSP 2007; key-value storage system
Amazon Dynamo

View Slide

Focused on high-availability and low-latency
Amazon Dynamo

View Slide

Collection of distributed systems techniques
Amazon Dynamo

View Slide

LinkedIn Voldemort, Facebook Cassandra
Amazon Dynamo

View Slide

Released 2009; Apache2 licensed Dynamo clone
Basho Riak

View Slide

Installing and Using Riak

View Slide

$ curl -O http://s3.amazonaws.com/downloads.basho.com/erlang/
otp_src_R16B02-basho5.tar.gz
$ tar -xvf otp_src_R16B02-basho5.tar.gz
$ cd otp_src_R16B02-basho5
$ ./configure && make && sudo make install
Installing Erlang

View Slide

$ git clone https://github.com/basho/riak.git
$ cd riak
$ make all
Building Riak

View Slide

$ make devrel DEVNODES=5
$ cd dev; ls
Building a devrel

View Slide

$ for node in dev*; do $node/bin/riak start; done
Starting a devrel

View Slide

$ for node in dev*; do $node/bin/riak ping; done
Pinging all nodes in a devrel

View Slide

$ dev2/bin/riak-admin cluster join [email protected]
Stage a join

View Slide

$ dev1/bin/riak-admin cluster plan
View a staged plan

View Slide

=============================== Staged Changes ================================
Action Nodes(s)
-------------------------------------------------------------------------------
join '[email protected]'
-------------------------------------------------------------------------------
NOTE: Applying these changes will result in 1 cluster transition
###############################################################################
After cluster transition 1/1
###############################################################################
================================= Membership ==================================
Status Ring Pending Node
-------------------------------------------------------------------------------
valid 100.0% 20.3% '[email protected]'
-------------------------------------------------------------------------------
Valid:5 / Leaving:0 / Exiting:0 / Joining:0 / Down:0
Transfers resulting from cluster changes: 51
12 transfers from '[email protected]' to '[email protected]'
View a staged plan

View Slide

$ dev2/bin/riak-admin cluster commit
Commit the plan

View Slide

$ dev1/bin/riak-admin member-status
View members of cluster

View Slide

================================= Membership ==================================
Status Ring Pending Node
-------------------------------------------------------------------------------
valid 20.3% -- '[email protected]'
-------------------------------------------------------------------------------
Valid:5 / Leaving:0 / Exiting:0 / Joining:0 / Down:0
View members of cluster

View Slide

$ curl -XPUT http://localhost:10018/buckets/welcome/keys/german -H
'Content-Type: text/plain' -d 'herzlich willkommen'
Storing data via HTTP

View Slide

$ curl http://localhost:10018/buckets/welcome/keys/german
Retrieving data via HTTP

View Slide

$ curl -XPUT http://localhost:10018/buckets/images/keys/
.jpg \
-H 'Content-Type: image/jpeg' \
--data-binary @.jpg
Storing an image via HTTP

View Slide

$ curl -O http://localhost:10018/buckets/images/keys/.jpg
Retrieving an image via HTTP

View Slide

Riak Architecture

View Slide

Consistent Hashing
hash(bucket/key)

View Slide

hash ring

View Slide

tokenize it

View Slide

node 0
node 1
node 2
hash(key)

View Slide

node 0
node 1
node 2
Replicas are stored to the N - 1
contiguous partitions

View Slide

node 0
node 1
node 2
hash(companies/cisco)
Replicas are stored to the N - 1
contiguous partitions

View Slide

node 0
node 1
node 2

View Slide

Scaling out
node 0
node 1
node 2
node 3 +

View Slide

Quorum
requests
N R W PR/PW DW

View Slide

Vector Clocks
establish temporality

View Slide

View Slide

Anatomy of a Request
get(users/clay-davis)

View Slide

client
Riak

View Slide

Get Handler (FSM)
client
Riak

View Slide

Get Handler (FSM)
client
Riak
hash(users/clay-davis)
== 10, 11, 12

View Slide

Get Handler (FSM)
client
Riak
hash(users/clay-davis)
== 10, 11, 12
Coordinating node
Cluster
6 7 8 9 10 11 12 13 14 15 16
The Ring

View Slide

Get Handler (FSM)
client
Riak
Coordinating node
Cluster
6 7 8 9 10 11 12 13 14 15 16
The Ring

View Slide

Get Handler (FSM)
client
Riak
Coordinating node
Cluster
6 7 8 9 10 11 12 13 14 15 16
The Ring
R=2

View Slide

Get Handler (FSM)
client
Riak
Coordinating node
Cluster
6 7 8 9 10 11 12 13 14 15 16
The Ring
R=2 obj

View Slide

Get Handler (FSM)
client
Riak
R=2 obj obj

View Slide

Get Handler (FSM)
client
Riak
R=2 obj
obj

View Slide

obj

View Slide

Read Repair
(Anti-Entropy)

View Slide

replica replica replica

View Slide

X

View Slide


View Slide

Active Anti-Entropy
(self healing clusters)

View Slide

real-time updates
persistent
non-blocking
disk-based

View Slide

merkle tree to track changes
coordinated at the vnode level
runs as a background process
exchange with
neighbor vnodes for inconsistencies
resolution semantics:
trigger read-repair

View Slide

= hashes marked dirty

View Slide

View Slide

= keys to read-repair

View Slide

Riak and Consistency

View Slide

Riak Object

View Slide

BKey Value

View Slide

Consistent hashing; dynamic membership
Data Placement

View Slide

View Slide

Replication per-value across ring
Data Placement

View Slide

Replica Replica Replica

View Slide

High Availability
…any non-failing node can respond to any request
Gilbert & Lynch

View Slide

Eventual Consistency
Eventual consistency is a consistency model used in
distributed computing that informally guarantees that, if no
new updates are made to a given data item, eventually all
accesses to that item will return the last updated value.
Wikipedia

View Slide

Take the form: {Writer, Value, Time}
Concurrent writes

View Slide

[{a, v1, t1}] [{b, v1, t2}] [{a, v1, t1}]
Concurrent writes

View Slide

[{a, v1, t1}] [{b, v1, t2}] [{a, v1, t1}]
[{b, v1, t2}] [{b, v1, t2}] [{b, v1, t2}]
Last Writer Wins

View Slide

[{a, v1, t1}] [{b, v1, t2}] [{a, v1, t1}]
[[{a, v1, t1}, {b, v1, t2}] [[{a, v1, t1}, {b, v1, t2}] [[{a, v1, t1}, {b, v1, t2}]
Allow Mult

View Slide

User specificed
Merge

View Slide

Two Approaches

View Slide

Strong
Eventual Consistency

View Slide

Designed for convergence; allows divergence
Conflict-free Replicated Data Types

View Slide

Strong
Consistency

View Slide

Provides atomicity and recency
Strong Consistency

View Slide

Prohibits partial writes
Strong Consistency

View Slide

A A A

View Slide

A A A
Val = B

View Slide

B A A

View Slide

B A A
Get Operation with Read Repair

View Slide

B A A
Get Operation with Read Repair
B B

View Slide

Single key atomic operations
Strong Consistency

View Slide

Requires read/modify/write cycle (CAS)
Strong Consistency

View Slide

Consensus

View Slide

Distributed Consensus
The problem of reaching agreement among remote
processes is one of the most fundamental problems in
distributed computing and is at the core of many
algorithms for distributed data processing,
distributed file management, and fault-tolerant
distributed applications.
Fischer, Lynch, Paterson

View Slide

Termination, agreement, validity
The Consensus Problem

View Slide

All processes eventually decide on a value
Termination

View Slide

All processes decide on the same value
Agreement

View Slide

Value decided on had to have been proposed
Validity

View Slide

Consensus Algorithms

View Slide

Paxos, ZAB, Raft, etc.
Consensus Algorithms

View Slide

Coordinated requests with a chosen leader
The Paxos Algorithm

View Slide

Node 1 Node 2 Node 3
N++ prepare(N)
promise(N, Vb) promise(N, Vc)
Vn = f(Va, Vb, Vc) commit(N, Vn)
accept(N)

View Slide

First request
Multi-Paxos

View Slide

N++; I = 0 prepare(N, I)
promise(N, I, Vb) promise(N, I, Vc)
Vn = f(Va, Vb, Vc) commit(N, I, Vn)
accept(N, I)

View Slide

Each additional request
Multi-Paxos

View Slide

I++ commit(N, I, V)
accept(N, I)

View Slide

Ship entire state!
Multi-Paxos

View Slide

Riak

View Slide

Key-value store; keys are independent state
Riak

View Slide

Multi-Paxos per key; CAS on isolated state
Riak

View Slide

Consensus Groups

View Slide

Participants in decisioning; ensembles
Consensus Groups

View Slide

Use the preference list!
Consensus Groups

View Slide

preflist

View Slide

View Slide

One ensemble per preference list; ring size
Consensus Groups

View Slide

Ensembles

View Slide

election of leader; get/put operations
Riak Ensembles

View Slide

read local; refresh, if old
Get Operations

View Slide

obj.epoch < epoch get(key)
reply(Epochb, Seqb, Valb)
Val = latest(Vala, Valb, Valc)
Val.epoch = epoch
write(Epoch, ++Seq, Val)
ack(Epoch, Seq)
reply(Epochc, Seqc, Valc)

View Slide

obj.epoch == epoch
Reply = local_get(Key)

View Slide

Worst Case: 2 roundtrips / write
Get Operations
Best Case: 0 roundtrips / write

View Slide

read local; refresh, modify and commit if old
Put Operations

View Slide

obj.epoch < epoch get(key)
reply(Epochb, Seqb, Valb)
Latest = latest(Vala, Valb, Valc)
Val = modify(Latest)
ack(Epoch, Seq)
reply(Epochc, Seqc, Valc)

View Slide

obj.epoch == epoch
Latest = local_get(Key)
Val = modify(Latest)
ack(Epoch, Seq)

View Slide

Worst Case: 2 roundtrips / write
Put Operations
Best Case: 1 roundtrips / write

View Slide

Elect a new leader; start a new epoch
Failed Quorums

View Slide

Cluster Membership

View Slide

Use joint consensus from multi paxos
Dynamic Membership

View Slide

Existing Ensemble Joining Ensemble
riak_01
riak_02
riak_03
riak_07
riak_08
riak_09
[{riak_01}, {riak_02}, {riak_03}] [{riak_07}, {riak_08}, {riak_09}]

View Slide

Joint-Consensus Ensemble
[{riak_01}, {riak_02}, {riak_03}, {riak_07}, {riak_08}, {riak_09}]

View Slide

New Ensemble
riak_07
riak_08
riak_09
[{riak_07}, {riak_08}, {riak_09}]

View Slide

Distributed batch processing for Riak
MapReduce

View Slide

Data locality for map; coordinator for reduce
MapReduce

View Slide

View Slide

$ curl -XPUT http://localhost:10018/buckets/training/keys/foo \
-H 'Content-Type: text/plain' \
-d 'caremad data goes here'
$ curl -XPUT http://localhost:10018/buckets/training/keys/bar \
-d 'caremad caremad caremad caremad'
$ curl -XPUT http://localhost:10018/buckets/training/keys/baz \
-d 'nothing to see here'
$ curl -XPUT http://localhost:10018/buckets/training/keys/bam \
-d 'caremad caremad caremad'
Create some objects; http://docs.basho.com/riak/latest/dev/using/mapreduce/

View Slide

> ReFun = fun(O, _, Re) ->
case re:run(riak_object:get_value(O), Re, [global]) of
{match, Matches} -> [{riak_object:key(O), length(Matches)}];
nomatch -> [{riak_object:key(O), 0}]
end
end.
> {ok, Re} = re:compile("caremad").
> {ok, Riak} = riakc_pb_socket:start_link("127.0.0.1", 8087).
> riakc_pb_socket:mapred_bucket(Riak, <<"training">>,
[{map, {qfun, ReFun}, Re, true}]).
Run Erlang MapReduce; http://docs.basho.com/riak/latest/dev/using/mapreduce/

View Slide

Distributed secondary indexing over values
Secondary Indexes (2i)

View Slide

Requires LevelDB or memory backend

View Slide

Tag objects; perform equality or range queries

View Slide

$ curl -XPOST localhost:8098/types/mytype/buckets/users/keys/
john_smith \
-H 'x-riak-index-twitter_bin: jsmith123' \
-H 'x-riak-index-email_bin: [email protected]' \
-H 'Content-Type: application/json' \
-d '{"userData":"data"}'
Create values with secondary index tags; http://docs.basho.com/riak/latest/dev/using/2i/

View Slide

$ curl http://localhost:10018/buckets/users/index/twitter_bin/
jsmith123
Query secondary index; http://docs.basho.com/riak/latest/dev/using/2i/

View Slide

Riak integration with Solr Distributed Search
Riak Search

View Slide

View Slide

Schemas explain how to index fields
Riak Search Components

View Slide

Indexes are built and queried against

View Slide

Bucket-Index associations say when to index

View Slide

Default schema covers many content-types

View Slide

$ curl -XPUT http://localhost:10018/search/index/famous
Create default index using default schema; http://docs.basho.com/riak/latest/dev/using/search/

View Slide

$ curl -XPUT http://localhost:10018/search/index/famous \
-d '{"schema":"_yz_default"}'
Create default index using default schema; http://docs.basho.com/riak/latest/dev/using/search/

View Slide

$ riak-admin bucket-type create animals '{"props":{}}'
$ riak-admin bucket-type activate animals
Create bucket type for search; http://docs.basho.com/riak/latest/dev/using/search/

View Slide

$ curl -XPUT http://localhost:10018/types/animals/buckets/cats/props \
-d '{"props":{"search_index":"famous"}}'
Associate bucket, bucket type, and index; http://docs.basho.com/riak/latest/dev/using/search/

View Slide

$ curl -XPUT http://localhost:10018/types/animals/buckets/cats/keys/liono \
-d '{"name_s":"Lion-o", "age_i":30, "leader_b":true}'
$ curl -XPUT http://localhost:10018/types/animals/buckets/cats/keys/cheetara \
-d '{"name_s":"Cheetara", "age_i":28, "leader_b":false}'
$ curl -XPUT http://localhost:10018/types/animals/buckets/cats/keys/snarf \
-d '{"name_s":"Snarf", "age_i":43}'
$ curl -XPUT http://localhost:10018/types/animals/buckets/cats/keys/panthro \
-d '{"name_s":"Panthro", "age_i":36}'
Store some values; http://docs.basho.com/riak/latest/dev/using/search/

View Slide

$ curl “http://localhost:10018/search/query/famous?
wt=json&q=name_s:Lion*” | jsonpp
$ curl “http://localhost:10018/search/query/famous?wt=json&q=age_i:
%5B30%20TO%20*%5D” | jsonpp
$ curl “http://localhost:10018/search/query/famous?
wt=json&q=leader_b:true%20AND%20age_i:%5B25%20TO%20*%5D” | jsonpp
Perform search queries; http://docs.basho.com/riak/latest/dev/using/search/

View Slide

Single-key linearizability; reduced availability
Strong Consistency

View Slide

$ riak-admin bucket-type create strongly_consistent \
‘{"props":{"consistent":true}}'
$ riak-admin bucket-type status strongly_consistent
$ riak-admin bucket-type activate strongly_consistent
Enable strong consistency; http://docs.basho.com/riak/latest/dev/advanced/strong-consistency/

View Slide

Read and write a value to SC bucket
Exercise

View Slide

Conflict-Free Replicated Data Types
Strong Eventual Consistency

View Slide

Converge correctly under concurrent ops *
Strong Eventual Consistency
* See the next talk from Annette Bieniusa!

View Slide

$ riak-admin bucket-type create maps \
'{"props":{"datatype":"map"}}'
$ riak-admin bucket-type create sets \
'{"props":{"datatype":"set"}}'
$ riak-admin bucket-type create counters \
‘{“props":{"datatype":"counter"}}'
$ riak-admin bucket-type status maps
$ riak-admin bucket-type activate maps
Create bucket type for data types; http://docs.basho.com/riak/latest/dev/using/data-types/

View Slide

$ curl -XPOST http://localhost:10018/types/counters/buckets/counters/
datatypes/traffic_tickets \
-H "Content-Type: application/json" \
-d '{"increment": 1}’
$ curl http://localhost:10018/types/counters/buckets/counters/
datatypes/traffic_tickets
Operate on counters; http://docs.basho.com/riak/latest/dev/using/data-types/

View Slide

$ curl -XPOST http://localhost:10018/types/sets/buckets/travel/
datatypes/cities \
-d '{"add_all":["Toronto", “Montreal"]}'
$ curl -XPOST http://localhost:10018/types/sets/buckets/travel/
datatypes/cities \
-d '{"remove": “Montreal"}'
$ curl http://localhost:10018/types/sets/buckets/travel/datatypes/
cities
Operate on sets; http://docs.basho.com/riak/latest/dev/using/data-types/

View Slide

$ curl -XPOST http://localhost:10018/types/maps/buckets/customers/
datatypes/ahmed_info \
-d '
{
"update": {
"first_name_register": "Ahmed",
"phone_number_register": "5551234567"
}
}'
$ curl -XPOST http://localhost:8098/types/maps/buckets/customers/
datatypes/ahmed_info \
-d '
{
"update": {
"annika_info_map": {
"update": {
"interests_set": {
"add": "tango dancing"
}
}
}
}
}
' Operate on maps; http://docs.basho.com/riak/latest/dev/using/data-types/

View Slide

Read and write a value to map
Exercise

View Slide

Questions?

View Slide

Introduction to Riak

Introduction to Riak

Christopher Meiklejohn

More Decks by Christopher Meiklejohn

Other Decks in Programming

Featured

Transcript