Riak add presentation

Riak — простая, предсказуемая и
масштабируемая БД.
Преимущества и недостатки.
Илья Богунов
Профессионалы.ру
twitter.com/tech_mind

О чем это ?
• Исторический экскурс БД от mainframe`ов 80х к вебу
2000-ных и текущим временам.
• Что такое Riak. Как он выглядит для
разработчикаадмина.
• Кто его и зачем использует.
• Дополнительные плюшки, кроме чистого kv.
• Не silver bullet - wtf`ы, проблемы и
моменты, которые надо учитывать.
• Код и коммьюнити.
• «Кто, где и когда» ? Зачем и где его использовать.

http://www.flickr.com/photos/lofink/4501610335/

DB: mainframe`ы
As an OLTP system benchmark, TPC-C simulates a complete
environment where a population of terminal operators
executes transactions against a database.
The benchmark is centered around the principal activities
(transactions) of an order-entry environment.
These transactions include entering and delivering
orders (insert, update),
recording payments (insert, update),
checking the status of orders (select), and
monitoring the level of stock at the warehouses (group
by, sum, etc).
http://www.tpc.org/tpcc/detail.asp

DB/APP 80-x: интерфейсы
(найдите 10 отличий)

DB: MAINFRAME -> WEB

60-70е + = SQL

80е
= SQL

= (My)SQL
90е-2000е
+

2000++

+ + = SQL ?

MySql
• Replica (v 3.23.15, 2000)
• memcached (2003)
• batch insert/queue (mainframe`s)
• c/java daemon (custom binary file format)
• sharding (vertical/horizontal) + hand-made
sharding framework


SHARDING

Middle layer
http://www.dbshards.com/dbshards/

SHARDING

EAV
http://backchannel.org/blog/friendfeed-schemaless-mysql Sharding framework
https://github.com/twitter/gizzard

mysql patching
https://www.facebook.com/note.php?note_id=332335025932

Dynamo: Amazon’s Highly Available Key-value Store (2007),
http://www.allthingsdistributed.com/files/amazon-dynamo-sosp2007.pdf

Amazon Dynamo Clones:
• Cassandra (2008)
• Voldemort (2009) DHT
• Riak (2009)

RIAK:FOR DEVELOPER
RIAK(KV) SQL
• INSERT/CREATE/UPDATE/DELETE/, CREATE/ALT
• PUT/GET/DELETE ER/DROP, GRANT/REVOKE/DENY, MERGE/JOI
N/GROUP BY/HAVING/UNION …
Client:get(<<“bucket”>>, <<“key”>>)

• A (req/seconds) • X (req/seconds)
× B (% file cache miss) × [index, locks, read-ahead, random-
≈ C iops reads, optimizator index
selection, cache, buffer sizes]
= WTF ???

RIAK:FOR DEVELOPER
RIAK(KV) vs SQL

Client:mapred(List, fun () -> doX() end) = length(List) io reqs
GROUP BY/COUNT = ? io (random read + filesort) reqs

pessimistic/optimistic locking, WRITE ALWAYS, FIX ON READ
START TRANSACTION/COMMIT/ROLLBACK

`cat /etc/riak/app.config | grep "s*{" | grep -v "%%" | wc -l` = 78
`mysql -u root -e "show variables" | wc -l` = 301

ALTER
ALTER DOWNTIMECUSTOM MIGRATION SCRIPT

RIAK: FOR ADMIN/OPS
• Read repair, failover, hinted handoff.
• горизонтальное масштабирование
• легкость установки и добавления
ноды:
apt-getpkgindpkg install riak
riak start
riak-admin join node@name

CASSANDRA
http://wiki.apache.org/cassandra/Operations#Bootstrap
• Bootstrap
• Adding new nodes is called "bootstrapping."
• To bootstrap a node, turn AutoBootstrap on in the configuration file, and start it.
• If you explicitly specify an InitialToken in the configuration, the new node will bootstrap to that position on the ring. Otherwise, it will pick a Token that will give it half the keys from the node with the most disk space used, that does not already have
another node bootstrapping into its Range.
• Important things to note:
• You should wait long enough for all the nodes in your cluster to become aware of the bootstrapping node via gossip before starting another bootstrap. The new node will log "Bootstrapping" when this is safe, 2 minutes after starting. (90s to make sure
it has accurate load information, and 30s waiting for other nodes to start sending it inserts happening in its to-be-assumed part of the token ring.)
• Relating to point 1, one can only bootstrap N nodes at a time with automatic token picking, where N is the size of the existing cluster. If you need to more than double the size of your cluster, you have to wait for the first N nodes to finish until your
cluster is size 2N before bootstrapping more nodes. So if your current cluster is 5 nodes and you want add 7 nodes, bootstrap 5 and let those finish before bootstrapping the last two.
• As a safety measure, Cassandra does not automatically remove data from nodes that "lose" part of their Token Range to a newly added node. Run nodetool cleanup on the source node(s) (neighboring nodes that shared the same subrange) when you
are satisfied the new node is up and working. If you do not do this the old data will still be counted against the load on that node and future bootstrap attempts at choosing a location will be thrown off.
• When bootstrapping a new node, existing nodes have to divide the key space before beginning replication. This can take awhile, so be patient.
• During bootstrap, a node will drop the Thrift port and will not be accessible from nodetool.
• Bootstrap can take many hours when a lot of data is involved. See Streaming for how to monitor progress.
• Cassandra is smart enough to transfer data from the nearest source node(s), if your EndpointSnitch is configured correctly. So, the new node doesn't need to be in the same datacenter as the primary replica for the Range it is bootstrapping into, as
long as another replica is in the datacenter with the new one.
• Bootstrap progress can be monitored using nodetool with the netstats argument (0.7 and later) or streams (Cassandra 0.6).
• During bootstrap nodetool may report that the new node is not receiving nor sending any streams, this is because the sending node will copy out locally the data they will send to the receiving one, which can be seen in the sending node through the
the "AntiCompacting... AntiCompacted" log messages.
• Moving or Removing nodes
• Removing nodes entirely
• You can take a node out of the cluster with nodetool decommission to a live node, or nodetool removetoken (to any other machine) to remove a dead one. This will assign the ranges the old node was responsible for to other nodes, and replicate the
appropriate data there. If decommission is used, the data will stream from the decommissioned node. Ifremovetoken is used, the data will stream from the remaining replicas.
• No data is removed automatically from the node being decommissioned, so if you want to put the node back into service at a different token on the ring, it should be removed manually.
• Moving nodes
• nodetool move: move the target node to a given Token. Moving is both a convenience over and more efficient than decommission + bootstrap. After moving a node, nodetool cleanup should be run to remove any unnecessary data.
• As with bootstrap, see Streaming for how to monitor progress.
• Load balancing
• If you add nodes to your cluster your ring will be unbalanced and only way to get perfect balance is to compute new tokens for every node and assign them to each node manually by using nodetool move command.
• Here's a python program which can be used to calculate new tokens for the nodes. There's more info on the subject at Ben Black's presentation at Cassandra Summit 2010. http://www.datastax.com/blog/slides-and-videos-cassandra-summit-2010
• def tokens(nodes):
– for x in xrange(nodes):
• print 2 ** 127 / nodes * x
• In versions of Cassandra 0.7.* and lower, there's also nodetool loadbalance: essentially a convenience over decommission + bootstrap, only instead of telling the target node where to move on the ring it will choose its location based on the same
heuristic as Token selection on bootstrap. You should not use this as it doesn't rebalance the entire ring.
• The status of move and balancing operations can be monitored using nodetool with the netstat argument. (Cassandra 0.6.* and lower use the streams argument).
• Replacing a Dead Node
• Since Cassandra 1.0 we can replace a dead node with a new one using the property "cassandra.replace_token=<Token>", This property can be set using -D option while starting cassandra demon process.
• (Note:This property will be taken into effect only when the node doesn't have any data in it, You might want to empty the data dir if you want to force the node replace.)
• You must use this property when replacing a dead node (If tried to replace an existing live node, the bootstrapping node will throw a Exception). The token used via this property must be part of the ring and the node have died due to various reasons.
• Once this Property is enabled the node starts in a hibernate state, during which all the other nodes will see this node to be down. The new node will now start to bootstrap the data from the rest of the nodes in the cluster (Main difference between
normal bootstrapping of a new node is that this new node will not accept any writes during this phase). Once the bootstrapping is complete the node will be marked "UP", we rely on the hinted handoff's for making this node consistent (Since we don't
accept writes since the start of the bootstrap).
• Note: We Strongly suggest to repair the node once the bootstrap is completed, because Hinted handoff is a "best effort and not a guarantee".

VOLDEMORT
https://github.com/voldemort/voldemort/wiki/Voldemort-Rebalancing

• What are the actual steps performed during rebalancing?
• Following are the steps that we go through to rebalance successfully. The controller initiates the rebalancing and then waits for the completion.
• Input
– Either (a) current cluster xml, current stores xml, target cluster xml (b) url, target cluster xml
– batch size – Number of primary partitions to move together. There is a trade-off, more primary partitions movements = more redirections.
– max parallel rebalancing – Number of units ( stealer + donor node tuples ) to move together
• Get the latest state from the cluster
Compare the targetCluster.xml provided and add all new nodes to currentCluster.xml
• Verify that
– We are not in rebalancing state already ( “./bin/voldemort-admin-tool.sh —get-metadata server.state —url [url+” all returns NORMAL_SERVER and “./bin/voldemort-admin-tool.sh —get-metadata rebalancing.steal.info.key —url [url+” all returns “*+” i.e. no rebalancing state )
– RO stores ( if they exist ) are all using the latest storage format ( “./bin/voldemort-admin-tool.sh —ro-metadata storage-format —url [url+” returns all stores with “ro2” format )
• Get a list of every primary partition to be moved
• For every “batch” of primary partitions to move
– Create a transition cluster metadata which contains movement of “batch size” number of primary partitions
– Create a rebalancing plan based on this transition cluster.xml and the current state of the cluster. The plan generated is a map of stealer node to list of donor node + partitions to be moved.
– State change step
• has_read-only_stores AND has_read-write_stores => Change the rebalancing state [ with RO stores information ] change on all stealer nodes
• has_read-only_stores => Change the rebalancing state change on all stealer nodes
– Start multiple units of migration [ unit => stealer + donor node movement ] of all RO Stores data [ No changes to the routing layer and clients ]. At the end of each migration delete the rebalancing state => Done with parallelism ( max parallel rebalancing )
– State change step [ Changes in rebalancing state will kick in redirecting stores which will start redirecting requests to the old nodes ]
• hasROStore AND hasRWStore => Change the cluster metadata + Swap on all nodes AND Change rebalancing state [ with RW stores information ] on all stealer nodes
• hasROStore AND !hasRWStore => Change the cluster metadata + Swap on all nodes
• !hasROStore AND hasRWStore => Change the cluster metadata on all nodes AND Change rebalancing state [ with RW stores information ] on all stealer nodes
– Start multiple units of migration of all RW Stores data [ With redirecting happening ]. At the end of migration delete the rebalancing state => Done with parallelism ( max parallel rebalancing )
• What about the failure scenarios?
• Extra precaution has been taken and every step ( 5 [ c,d,e,f ] ) has a rollback strategy
• Rollback strategy for
5 ( c , e ) => If any failure takes place during the ‘State change step’, the following rollback strategy is run on every node that was completed successfully
• Swap ROChange cluster metadataChange rebalance stateOrderFTTremove the rebalance state change → change back clusterFFTremove from rebalance state changeTTFchange back cluster metadata → swapTTTremove from rebalance state change → change back cluster → swap5 ( d, f ) => Similarly during migration of partitions
• Has RO storesHas RW storesFinished RO storesRollback Action [ in the correct order ]TTTrollback cluster change + swapTTFnothing to do since “rebalance state change” should have removed everythingTFTwon’t be triggered since hasRW is falseTFFnothing to do since “rebalance state change” should have removed everythingFTTrollback cluster changeFTFwon’t be triggeredFFTwon’t be triggeredFFFwon’t be triggeredWhat happens on the stealer node side during 5 ( d, f )?
• The stealer node on receiving a “unit of migration” * unit => single stealer + single donor node migration + and does the following
• Check if the rebalancing state change was already done [ i.e. 5 ( c, e ) was successfully completed ]
• Acquire a lock for the donor node [ Fail if donor node was already rebalancing ]
• Start migration of the store partitions from the donor node => PARALLEL [ setMaxParallelStoresRebalancing ]. At the end of every store migration remove it from the list rebalance state change [ so as to stop redirecting stores ]
• What about the donor side?
• The donor node has no knowledge for rebalancing at all and keeps behaving normally.
• What about my data consistency during rebalancing?
• Rebalancing process has to maintain data consistency guarantees during rebalancing, We are doing it through a proxy based design. Once rebalancing starts the stealer node is the new master for all rebalancing partitions, All the clients talk directly to stealer node for all the requests for these partitions. Stealer node internally make proxy calls to the original donor node to return correct data back to the client.The process steps are
• Client request stealer node for key ‘k1’ belonging to partition ‘p1’ which is currently being migrated/rebalanced.
• Stealer node looks at the key and understands that this key is part of a rebalancing partition
• Stealer node makes a proxy call to donor node and gets the list of values as returned by the donor node.
• Stealer node does local put for all (key,value) pairs ignoring all ObsoleteVersionException
• Stealer node now should have all the versions from the original node and now does normal local get/put/ getAll operations.
• And how do my clients know the changes?
• Voldemort client currently bootstrap from the bootstrap URL at the start time and use the returned cluster/stores metadata for all subsequent operation. Rebalancing results in the cluster metadata change and so we need a mechanism to tell clients that they should rebootstrap if they have old metadata.
• Client Side routing bootstrapping: Since the client will be using the old metadata during rebalancing the server now throws anInvalidMetadataException if it sees a request for a partition which does not belong to it. On seeing this special exception the client is re-bootstraps from the bootstrap url and will hence pick up the correct cluster metadata.
• Server side routing bootstrapping: The other method of routing i.e. make calls to any server with enable_routing flag set to true and with re-routing to the correct location taking place on the server side ( i.e. 2 hops ). In this approach we’ve added a RebootstrappingStore which picks up the new metadata in case of change.
• How to start rebalancing?
• Step 1: Make sure cluster is not doing rebalancing.
– The rebalancing steal info should be null
./bin/voldemort-admin-tool.sh —get-metadata rebalancing.steal.info.key —url [ url ]
– The servers should be NORMAL_STATE
./bin/voldemort-admin-tool.sh —get-metadata server.state —url [ url ]
• To check whether the keys were moved correctly we need to save some keys and later check if they have been migrated to their new locations
./bin/voldemort-rebalance.sh —current-cluster [ current_cluster ] —current-stores [ current_stores_path ] —entropy false —output-dir * directory where we’ll store the keys for later use. Keys are stored on a per store basis +
• Generate the new cluster xml. This can be done either by hand or if you want to do it automatically here is a tool to do to
./bin/voldemort-rebalance.sh —current-cluster [ current_cluster_path ] —target-cluster [ should be the same as current-cluster but with new nodes put in with empty partitions ] —current-stores [ current_stores_path ] —generate
• Run the new metadata through key-distribution generator to get an idea of skew if any. Make sure your standard deviation is close to 0.
./bin/run-class.sh voldemort.utils.KeyDistributionGenerator —cluster-xml [ new_cluster_xml_generated_above ] —stores-xml [ stores_metadata ]
• Use the new_cluster_xml and run the real rebalancing BUT first with —show-plan ( to check what we’re going to move )
./bin/voldemort-rebalance.sh —url [ url ] —target-cluster [ new_cluster_metadata_generated ] —show-plan
• Run the real deal
./bin/voldemort-rebalance.sh —url [ url ] —target-cluster [ new_cluster_metadata_generated ]
• Monitor by checking the async jobs
./bin/voldemort-admin-tool.sh —async get —url [ url ]
• The following takes place when we are running the real rebalancing (i) Pick batch of partitions to move (ii) Generate transition plan (iii) Execute the plan as a series of ‘stealer-donor’ node copying steps. By default we do only one ‘stealer-donor’ tuple movement at once. You can increase this by setting —parallelism option.
• If anything fails we can rollback easily ( as long as —delete was not used while running voldemort-rebalance.sh )
• To stop an async process
./bin/voldemort-admin-tool.sh —async stop —async-id [comma separated list of async jobs ] —url [ url ] —node [ node on which the async job is running ]
• To clear the rebalancing information on a particular node
./bin/voldemort-admin-tool.sh —clear-rebalancing-metadata —url [ url ] —node [ node-id ]
• Following are some of the configurations that will help you on the server side
• Parameter on serverDefaultWhat it doesenable.rebalancingtrueShould be true so as to run rebalancingmax.rebalancing.attempts3Once a stealer node receives the plan to copy from a donor node, it will attempt this many times to copy the data ( in case of failure )rebalancing.timeout.seconds10 * 24 * 60 * 60Time we give for the server side rebalancing to finish copying data from a donor nodemax.parallel.stores.rebalancing3Stores to rebalance in parallelrebalancing.optimizationtrueSome times
we have data stored without being partition aware ( Example : BDB ). In this scenario we can run an optimization phase which ignores copying data over if a replica already existsWhat is left to make this process better?
• a) Execute tasks should be smarter and choose tasks to execute so as to avoid two disk sweeps happening on the same node.
• b) Fix deletes! – Make it run at the end instead of in the middle * Even though I’ll never run this in production +
• c) Logging – Currently we propagate the message at the lowest level all the way to the top. Instead we should try to make a better progress bar ( “Number of stores completed – 5 / 10” ) and push that upwards.
• d) Currently the stealer node goes into REBALANCING_MASTER state and doesn’t allow any disk sweeps ( like slop pusher job, etc ) from not taking place. But what about the poor donor node

HAND-MADE SHARDING
apt-get install x-sql
./configure
Мигрируем
• легкий способ:
1. стопнуть сервис
2. запустить скрипты миграции
3. запустить сервис
• сложный способ:
1. дописать код работающий с 2мя партициями одновременно
2. научить приложение понимать, когда идет миграция и что надо
читатьписать в ОБА места
3. следить чтоб при переезде не разбалансировался кластер
4. ощутить после переезда всю прелесть неучтенных
многопоточных ситуацийнеконсистентных данных

RIAK: USES
• http://wiki.basho.com/Who-is-Using-Riak.html
• Mozilla http://blog.mozilla.com/data/2010/05/18/riak-and-cassandra-and-hbase-oh-my/
• Unitypark http://muchdifferent.com/
• Yammer http://dl.dropbox.com/u/2744222/2011-03-22_Riak-At-Yammer.pdf
• Voxer http://speakerd.s3.amazonaws.com/presentations/4f229a9421e6f8002201fa9f/ScaleConf-HealthyDistributedSystems.pdf

• Unison http://www.unison.com/
• Mochimedia http://www.mochimedia.com/
• Seomoz http://devblog.seomoz.org/2011/10/using-riak-for-ranking-collection/
• Trifork http://www.erlang-solutions.com/upload/docs/116/Basho%20and%20Trifork.pdf
• clipboard.com http://blog.clipboard.com/2012/03/18/0-Milking-Performance-From-Riak-Search
• Echo http://aboutecho.com/
• Другие =)

RIAK: КогдаГде ?
1. мне критична высоконадежность и
высокодоступность
2. у меня миллионы
юзеровприложенийстраниц
топиковфайловобьектовлент новостей
3. они между собой не связаны (нет real-time
GROUP BY/JOIN)
4. можно для аналитики типа time-series data
5. можно слепить ежа с ужом и прикрутить Riak
core например к Lucene

Riak: дополнительные плюшки
Индексы

Key = <<"X">>, Value = <<"X">>, IndexInt = 1.

Robj = riak_object:new(<<"X">>, Key, Value),

MetaData =
dict:from_list([{<<"index">>, [{<<"index_bin">>, <<IndexInt:32/
big-unsigned-integer>>}]}]),

Robj2 = riak_object:update_metadata(Robj, MetaData),

Client:put(Robj2, 1).

Client:get_index(<<"X">>, {range, <<"index_bin">>, <<1:32/big-
unsigned-integer>>, <<2:32/big-unsigned-integer>>}).

{ok,[<<"X">>]}

Полнотекстовый поиск

search:index_doc(<<"users">>, <<"test">>,
[{<<"title">>, <<"The Test">>}, {<<"content">>, <<"The
Content">>}]).

search:search(<<"users">>, <<"title:Test">>).

{1,
[{<<"users">>,<<"test">>,
[{p,[1]},{score,0.35355339059327373}]}]

• Postpre-commit хуки:
• Поиск – это тот же пре-коммит хук
• Интеграция с RabbitMq
https://github.com/jbrisbin/riak-rabbitmq-commit-hooks

• Map-reduce.
• На эрланге =)
• На javascript`e.
• По заданному списку ключей!

RIAK: (putgetdelete, uniform, 3vs6)

http://joyeur.com/2010/10/31/riak-smartmachine-benchmark-the-technical-details/

RIAK: (getrange_index, uniform, 6vs9)

RIAK: (puts, 3 nodes, node failure)

RIAK: GET vs MR-GET

mr-get(x100) Get(x1)

RIAK: Неудобства и WTF`S
Неудобства:
• Adhoc запросы невозможны, нужно ЗАРАНЕЕ
подумать о своих паттернах доступа
• COUNTGROUP BY, только пред-агрегацией
(map-reduce`ом довольно не удобно), либо
хранить счетчики в Redis`е
• Eventual consistency, планируем свои данные
чтоб уметь мерджить
• Нет паджинации и сортировки - либо
редис, либо что-то вроде linked-list`a

RIAK: Неудобства и WTF`S
WTF’S:
• BUCKET = PREFIX
• «урезанный» map-reduce
• Восстановление данных только по read-
repair на чтении, данные сами не починятся
(без EDS)
• RiakSearch – нет анти-энтропии
• Специфичные проблемы, каждого из
бекэндов

RIAK: WTF`S

Вывод - читаем mailing list и код =)

RIAK: CODE & COMMUNITY
https://github.com/basho/riak_kv/blob/master/src/riak_client.erl

https://github.com/basho/riak_kv/blob/master/src/riak_kv_get_fsm.erl


..........

https://github.com/basho/riak_kv/blob/master/src/riak_kv_vnode.erl

https://github.com/basho/riak_core/blob/master/src/riak_core_vnode_master.erl

https://github.com/basho/riak_kv/blob/master/src/riak_kv_eleveldb_backend.erl

Сообщество живое, иногда даже излишне)

• За ним стоит коммерческая компания
(с инвестициями), и в ней Eric Brewer =)
• Еженедельные дайджесты «что нового»

RIAK: Стоит или не стоит ?
• У вас есть проблема с количеством IOPS.
• Вам важна высокодоступность и хочется минимум
проблем при расширение кластера
• Вы готовы заранее полностью продумать то, как
ваши данные положить на kv
• Вы знаете, как будете решать конфликты
• Вам не нужны транзакции (pessimisticoptimistic
locking)
• Вы не боитесь erlang`a =) И не боитесь обратиться в
community, чтобы получить ответ на свой вопрос

RIAK: Стоит или не стоит ?

Спасибо за вниманиеВопросы.
Илья Богунов

bogunov@gmail.com
twitter.com/tech_mind

Пожалуйста, поставьте
оценку моему докладу.

Ваше мнение очень важно.

Спасибо!

Riak add presentation

Recommended

More Related Content

What's hot (20)

Viewers also liked (12)

Similar to Riak add presentation (20)

Recently uploaded (20)

Riak add presentation